Advanced Computing in Industrial Mathematics

13th Annual Meeting of the Bulgarian Section of

SIAM December 18 20 2018 Sofia Bulgaria Revised
1st Edition Ivan Georgiev download
https://textbookfull.com/product/advanced-computing-in-
industrial-mathematics-13th-annual-meeting-of-the-bulgarian-
section-of-siam-december-18-20-2018-sofia-bulgaria-revised-1st-
edition-ivan-georgiev/

Download full version ebook from https://textbookfull.com

We believe these products will be a great fit for you. Click
the link to download now, or visit textbookfull.com
to discover even more!

Advanced Computing in Industrial Mathematics Revised

Selected Papers of the 10th Annual Meeting of the
Bulgarian Section of SIAM December 21 22 2015 Sofia
Bulgaria 1st Edition Krassimir Georgiev
https://textbookfull.com/product/advanced-computing-in-
industrial-mathematics-revised-selected-papers-of-the-10th-
annual-meeting-of-the-bulgarian-section-of-siam-
december-21-22-2015-sofia-bulgaria-1st-edition-krassimir-
georgiev/
Soft Computing Systems Second International Conference
ICSCS 2018 Kollam India April 19 20 2018 Revised
Selected Papers Ivan Zelinka

https://textbookfull.com/product/soft-computing-systems-second-
international-conference-icscs-2018-kollam-india-
april-19-20-2018-revised-selected-papers-ivan-zelinka/

Advances in High Performance Computing: Results of the

International Conference on “High Performance
Computing” Borovets, Bulgaria, 2019 Ivan Dimov

https://textbookfull.com/product/advances-in-high-performance-
computing-results-of-the-international-conference-on-high-
performance-computing-borovets-bulgaria-2019-ivan-dimov/

ICT Education 47th Annual Conference of the Southern

African Computer Lecturers Association SACLA 2018
Gordon s Bay South Africa June 18 20 2018 Revised
Selected Papers Salah Kabanda
https://textbookfull.com/product/ict-education-47th-annual-
conference-of-the-southern-african-computer-lecturers-
association-sacla-2018-gordon-s-bay-south-africa-
Research in History and Philosophy of Mathematics The
CSHPM 2017 Annual Meeting in Toronto Ontario Maria Zack

https://textbookfull.com/product/research-in-history-and-
philosophy-of-mathematics-the-cshpm-2017-annual-meeting-in-
toronto-ontario-maria-zack/

Research in History and Philosophy of Mathematics The

CSHPM 2015 Annual Meeting in Washington D C 1st
Edition Maria Zack

https://textbookfull.com/product/research-in-history-and-
philosophy-of-mathematics-the-cshpm-2015-annual-meeting-in-
washington-d-c-1st-edition-maria-zack/

World Women in Mathematics 2018 Proceedings of the

First World Meeting for Women in Mathematics WM ²
Carolina Araujo

https://textbookfull.com/product/world-women-in-
mathematics-2018-proceedings-of-the-first-world-meeting-for-
women-in-mathematics-wm-%c2%b2-carolina-araujo/

Quantitative Psychology 83rd Annual Meeting of the

Psychometric Society New York NY 2018 Marie Wiberg

https://textbookfull.com/product/quantitative-psychology-83rd-
annual-meeting-of-the-psychometric-society-new-york-
ny-2018-marie-wiberg/

Large Scale Scientific Computing 11th International

Conference LSSC 2017 Sozopol Bulgaria June 5 9 2017
Revised Selected Papers 1st Edition Ivan Lirkov

https://textbookfull.com/product/large-scale-scientific-
computing-11th-international-conference-lssc-2017-sozopol-
bulgaria-june-5-9-2017-revised-selected-papers-1st-edition-ivan-
Studies in Computational Intelligence 961

Ivan Georgiev
Hristo Kostadinov
Elena Lilkova Editors

Advanced
Computing
in Industrial
Mathematics
13th Annual Meeting of the Bulgarian
Section of SIAM, December 18–20, 2018,
Sofia, Bulgaria, Revised Selected Papers
Studies in Computational Intelligence

Volume 961

Series Editor
Janusz Kacprzyk, Polish Academy of Sciences, Warsaw, Poland
The series “Studies in Computational Intelligence” (SCI) publishes new develop-
ments and advances in the various areas of computational intelligence—quickly and
with a high quality. The intent is to cover the theory, applications, and design
methods of computational intelligence, as embedded in the fields of engineering,
computer science, physics and life sciences, as well as the methodologies behind
them. The series contains monographs, lecture notes and edited volumes in
computational intelligence spanning the areas of neural networks, connectionist
systems, genetic algorithms, evolutionary computation, artificial intelligence,
cellular automata, self-organizing systems, soft computing, fuzzy systems, and
hybrid intelligent systems. Of particular value to both the contributors and the
readership are the short publication timeframe and the world-wide distribution,
which enable both wide and rapid dissemination of research output.
Indexed by SCOPUS, DBLP, WTI Frankfurt eG, zbMATH, SCImago.
All books published in the series are submitted for consideration in Web of
Science.

More information about this series at http://www.springer.com/series/7092

Ivan Georgiev Hristo Kostadinov
• •

Elena Lilkova
Editors

Advanced Computing
in Industrial Mathematics
13th Annual Meeting of the Bulgarian
Section of SIAM, December 18–20, 2018,
Sofia, Bulgaria, Revised Selected Papers

123
Editors
Ivan Georgiev Hristo Kostadinov
Institute of Information and Communication Institute of Mathematics and Informatics
Technologies and Institute of Mathematics Bulgarian Academy of Science
and Informatics Sofia, Bulgaria
Bulgarian Academy of Sciences
Sofia, Bulgaria

Elena Lilkova
Institute of Information
and Communication Technologies
Bulgarian Academy of Science
Sofia, Bulgaria

ISSN 1860-949X ISSN 1860-9503 (electronic)

Studies in Computational Intelligence
ISBN 978-3-030-71615-8 ISBN 978-3-030-71616-5 (eBook)
https://doi.org/10.1007/978-3-030-71616-5
© The Editor(s) (if applicable) and The Author(s), under exclusive license
to Springer Nature Switzerland AG 2021
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with regard
to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

The 13th Annual Meeting of the Bulgarian Section of the Society for Industrial and
Applied Mathematics (BGSIAM) was held in Sofia, December 18–20, 2018. The
section was formed in 2007 with the purpose to promote and support the application
of mathematics to science, engineering, and technology in Bulgaria.
The goals of BGSIAM follow and creatively develop the general goals of SIAM
as follows:
• To advance the application of mathematics and computational science to
engineering, industry, science, and society;
• To promote research that will lead to effective new mathematical and compu-
tational methods and techniques for science, engineering, industry, and society;
• To provide media for the exchange of information and ideas among mathe-
maticians, engineers, and scientists.
During the BGSIAM’18 conference, a wide range of problems concerning recent
achievements in the field of industrial and applied mathematics were presented and
discussed. The meeting provided a forum for exchange of ideas between scientists,
who develop and study mathematical methods and algorithms, and researchers, who
apply them for solving real-life problems.
Among the topics of interest are mathematical physics, numerical analysis,
high-performance computing, optimization and control, mathematical biology
stochastic modeling, machine learning, digitization and imaging, big data, and
advanced computing in environmental, biomedical, and engineering applications.
The invited speakers were:
• Owe Axelsson (Czech Academy of Sciences), An inner-product free iteration
method for an equation of motion and a bound-constrained optimal control PDE
problem
• Mauro Ballicchia (Polytechnic University of Marche, Italy), Recent advances in
the Wigner signed particles: theory and applications
• Vesselin Iossifov (University of Applied Sciences Berlin, Germany),
Programming Techniques for Energy-efficient Software

v
vi Preface

• Raytcho Lazarov (Texas A&M University, USA), Recent advances in numerical

methods for fractional differential equations with non-smooth data: a concise
overview
• Blagovest Sendov (Bulgarian Academy of Sciences), Algorithmic problems in
the geometry of polynomials
We would like to thank all the referees for the constructive remarks and criti-
cism, which furthered considerable improvements of the quality of the papers in this
book.
Contents

Design of a Multi-objective Optimization Model for Wireless

Sensor Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Alexander Alexandrov and Vladimir Monov
Cross-Validated Sequentially Constructed Multiple Regression
with Centered Predictors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Slav Angelov and Eugenia Stoimenova
Non-local Nonlinear Perturbation Analysis for a Nonlinear Matrix
Equation Arising in Tree-Like Stochastic Processes . . . . . . . . . . . . . . . . 16
Vera Angelova
One Parameter Family of Elliptic Curves and the Equation
x4 þ y4 þ kx2 y2 ¼ z2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Stoyan Apostolov, Miroslav Stoenchev, and Venelin Todorov
An Inner Product Free Solution Method for an Equation of Motion
with Indefinite Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Owe Axelsson
Simulation Model for Investigation of the Behaviour of an Artificial
Immune Algorithm for Generation of S-Boxes with Good
Cryptographic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Tsonka Baicheva, Violeta Andreeva, and Nikolay Nikolov
Self-consistent Monte Carlo Solution of Wigner and Poisson
Equations Using an Efficient Multigrid Approach . . . . . . . . . . . . . . . . . 60
Majid Benam, Maciej Wołoszyn, and Siegfried Selberherr
Modelled Versus Satellite Retrieved Estimation of the Direct Normal
Irradiance and the Sunshine Duration over Bulgaria . . . . . . . . . . . . . . . 68
Hristo Chervenkov and Kiril Slavov

vii
viii Contents

Numerical Identification of the Time Dependent Vertical Diffusion

Coefficient in a Model of Air Pollution . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Ivan T. Dimov, Juri D. Kandilarov, and Lubin G. Vulkov
Generalized Net Model for Flying Ant Colony Optimization . . . . . . . . . 90
Stefka Fidanova and Krassimir Atanassov
Numerical Determination of Time-Dependent Implied Volatility
by a Point Observation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Slavi G. Georgiev and Lubin G. Vulkov
Homotopy Analysis Method to Solve Volterra-Fredholm Fuzzy
Integral Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Atanaska Georgieva, Albena Pavlova, and Lozanka Trenkova
Spectral Fractional Laplacian with Inhomogeneous Dirichlet Data:
Questions, Problems, Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
S. Harizanov, S. Margenov, and N. Popivanov
Spike Timing Neural Model of Eye Movement Motor Response
with Reinforcement Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Petia Koprinkova-Hristova and Nadejda Bocheva
Programming Techniques for Energy-Efficient Software . . . . . . . . . . . . 154
Vesselin Iossifov
Modelling Human Biometeorological Conditions Using
Meteorological Data from Reanalysis and Objective
Analysis—Preliminary Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Vladimir Ivanov and Hristo Chervenkov
The Braided Group of a Square-Free Solution of the Yang-Baxter
Equation and Its Group Algebra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Tatiana Gateva-Ivanova
Exploring Applicability of the InterCriteria Analysis to Evaluate
the Performance of MOE and GOLD Scoring Functions . . . . . . . . . . . . 198
Dessislava Jereva, Tania Pencheva, Ivanka Tsakovska, Petko Alov,
and Ilza Pajeva
Multiobjective Approach for Solving Engineering Robust
Design Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Leoneed Kirilov and Petar Georgiev
Iterative Implicit Schemes for Compaction-Driven Darcy Flow
of Magma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
Miglena N. Koleva and Lubin G. Vulkov
Finite Time Blow up of the Solutions to Double Dispersive
Equations with Linear Restoring Force and High Energy . . . . . . . . . . . 233
N. Kolkovska, M. Dimova, and N. Kutev
Contents ix

Coupled Thermo-Hydro-Mechanical Finite Element Simulation

of Loess Slope Under Atmospheric Boundary Conditions . . . . . . . . . . . 244
Lingyun Li, Maria Datcheva, and Wiebke Baille
Molecular Dynamics Simulations of His6 -FLAG-hIFNc
Fusion Glycoproteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
E. Lilkova, N. Ilieva, P. Petkov, E. Krachmarova, G. Nacheva,
and L. Litov
Several Results Concerning the Barnes G-function, a Cosecant
Integral, and Some Other Special Functions . . . . . . . . . . . . . . . . . . . . . 268
Lubomir Markov
Influence of the Temperature on Simulated Annealing Method
for Metal Nanoparticle Structures Optimization . . . . . . . . . . . . . . . . . . 278
Rossen Mikhov, Vladimir Myasnichenko, Stefka Fidanova,
Leoneed Kirilov, and Nickolay Sdobnyakov
Blood Flow Instabilities Related to Elasticity and Asymmetry
of an Abdominal Aorta Aneurysm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
N. Nikolov, S. Tabakova, and St. Radev
Rank Data Clustering Based on Lee Distance . . . . . . . . . . . . . . . . . . . . 303
Nikolay I. Nikolov and Eugenia Stoimenova
Analysis of Swing Oscillatory Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Svetoslav G. Nikolov, Vassil M. Vassilev, and Daniela T. Zaharieva
Non-linear Waves of Interacting Populations
with Density-Dependent Diffusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
Elena V. Nikolova, Denislav Z. Serbezov, Ivan P. Jordanov,
and Nikolay K. Vitanov
Performance Analysis of Hierarchical Semi-separable Compression
Solver for Fractional Diffusion Problems . . . . . . . . . . . . . . . . . . . . . . . . 333
D. Slavchev
On the Spectral Properties of Lax Operators Related to BD.I
Symmetric Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
Alina Streche-Pauna, Aurelia Daniela Florian, and Vladimir S. Gerdjikov
Mechanical Properties Assessments for Materials of High Porosity
and Light Alloys with Predominant Embedded Phases . . . . . . . . . . . . . 359
Ludmila Parashkevova, Ludmil Drenchev, and Pedro Egizabal
Symmetries and Conservation Laws of a System of Timoshenko
Beam Type Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
Svilen I. Popov and Vassil M. Vassilev
x Contents

InterCriteria Analysis of the Human Factor Assessment

in a Mobile Company . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
Velichka Traneva and Stoyan Tranev
Using DLT in Software Lifecycle Management . . . . . . . . . . . . . . . . . . . 393
Biser Tsvetkov and Hristo Kostadinov
Study of a Flow of Reacting Substances in a Channel of Network . . . . . 405
Nikolay K. Vitanov, Kaloyan N. Vitanov, and Zlatinka I. Dimitrova
A Symplectic Numerical Method for the Sixth Order
Boussinesq Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
Veselina Vucheva and Natalia Kolkovska

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429

 Design of a Multi-objective Optimization
Model for Wireless Sensor Networks

Alexander Alexandrov(B) and Vladimir Monov

Institute of Information and Communication Technologies,

Bulgarian Academy of Sciences, Acad. G. Bonchev bl. 2, Sofia, Bulgaria
{akalexandrov,vmonov}@iit.bas.bg

Abstract. Wireless sensor networks (WSNs) have a worldwide influ-

ence during the latest years in the area of environmental monitoring.
This work addresses actual multi-objective optimization problems of the
wireless sensor networks used in this area. A new mathematical model
and multicriteria optimization approaches are presented in the paper
by taking into account different performance and shelf life objectives.
The proposed model is aimed at developing an energy-efficient commu-
nication between sensor nodes and the cluster heads (CH) in the wireless
sensor networks. We consider a few types of constraints concerning multi-
objective optimization of WSN with fixed place sensor nodes.

1 Introduction

Wireless sensor networks (WSNs) have had worldwide attention in the last years,
especially with the development of smart sensors and IoT technology. The new
smart sensors are small, and require less processing and computing resources,
and are relatively inexpensive. The smart sensors grouped in clusters can sense,
measure, integrate the measured data and transmit the sensed data to the con-
trol center. The trend towards wireless communication and IoT technology is
increasingly changing electronic devices in every aspect of life [1,3,19]. A WSN
is a special type of wireless network composed of a number of nodes that are
equipped with different sensors. This network type has been supported by tech-
nological advances in low-power wireless communications. A WSN is built by
distributed autonomous sensors that cooperatively monitor the environmental
conditions. To reduce the WSN design cost very offten are used modeling tools
as Generalized Nets (GN) [20] or industrial modelers as OPNET. In addition to
one or more sensors, each node in a sensor network is equipped with a micro-
controller, program logic, radio transceiver or other wireless communications
devices, and an energy source, usually based on battery. The size of the sensor
nodes results in corresponding resources such as energy, memory, computational
speed, and bandwidth [2]. In general, a WSN is composed of hundreds or even
thousands of small, low cost, low power wireless sensor nodes. These nodes are
deployed to register physical phenomena and usually communicate with each

c The Author(s), under exclusive license to Springer Nature Switzerland AG 2021

I. Georgiev et al. (Eds.): BGSIAM 2018, SCI 961, pp. 1–9, 2021.
https://doi.org/10.1007/978-3-030-71616-5_1
2 A. Alexandrov and V. Monov

other through wireless communication channels. The design of the most wireless
sensor networks is based mainly on star or cluster topology [4] that organizes
and maintains a group of wireless sensors with local integration data capability.
In the current research is considered a cluster based topology of WSN. Although
the existing wireless network technologies are capable of constructing of large
sensor networks, the design and use of sensor networks to monitor stationary or
mobile objects such as buildings, cars, and even personal hand-held devices can
be further simplified to reduce power consumption and to increase the device’s
shelf life. The simplest approach for communication in wireless sensor networks
is the direct transmission, in which each node transmits its own data directly to
the control center or to the cluster head. If the base station is far away, the cost
of sending data directly to that station will become too large, and the nodes will
consume their energy resources quickly. WSNs are now used in many industrial
and civilian application areas, including industrial process monitoring and con-
trol, machine health monitoring, environment, and habitat monitoring, health-
care applications, home automation, traffic control, and scientific exploration in
dangerous environments [10,16]. In all cases, however, the communication effi-
ciency and energy saving are one of the most important issues in the network.
The communication protocol specifies how wireless sensor nodes communicate
with each other, disseminating information that enables them to select the appro-
priate communication route between any two nodes in the WSN [17]. The choice
of route is executed via routing algorithms. Each router has a priori knowledge
of only the networks that are directly attached to it. A routing protocol shares
this information first with its immediate neighbors and then throughout the
network [18]. In this way, the routers can gain knowledge of the topology of the
network. Routing protocols designed for data transfer in traditional ad hoc net-
works cannot be used with sensor networks. This is because sensor nodes have
limited battery power and because data transmission or reception consumes more
power consumption than sensor nodes do sensing and computation operations.
Therefore, it is desirable to conserve the energy of the nodes in the network while
routing query responses back to the sink node.

2 Related Works

Due to the specific problems in multihop based WSN, the design of the com-
munication protocols poses many new challenges, and a lot of work has been
done in this area. During the last years, several papers have proposed more
and more efficient algorithms for wireless sensor based communication proto-
cols [5–7]. For example [6] proposes an adaptive routing protocol for fixed place
applications, using their IP addressing scheme in WSNs. The research of [7] pro-
poses an analytical model for a reliable real-time degree in multihop WSNs. The
real-time degree represents the amount of data that can be transmitted from
any source node to its destination node. The research of Al-Karaki [9] presents a
 Design of a Multi-objective Optimization Model 3

cluster-based algorithm addressing the communication problem under network

aggregation without reducing the data quality in WSNs. The paper is focused
on the problem of data exchange with data aggregation and maximizing the
sensor nodes lifetime via data encapsulation and different processing techniques.
The paper of Y.H.Zhua [14] proposes an approximation method for quality of
service (QoS) in WSN. The method uses Markov chains for two-dimensional
space merging. An algorithm using a virtual node infrastructure of WSNs is
presented in [11]. A probabilistic vector machine-based method in WSNs is pre-
sented in the research of Samadian [12]. An energy coverage control in WSNs
based on multi-objective genetic algorithm is presented in [15]. A mathemati-
cal model for energy efficient routing in WSN’s is presented in [13]. The paper
proposes a solution for the shortest path problem (SPP) and presents a model
with resource restrictions. At the same time, the proposed model is very general
and attempts to minimize only the sum of total WSN nodes energy by the end
of all transmission periods without considering the network topology and link
quality. Nevertheless, the considerations on this paper have potential as a basis
and initial point for further researches in the area of development of new commu-
nication protocols with integrated multi-objective optimization. The main focus
is adding new constraints and parameters related to network topology, adaptive
transmission periods related to inter-node communications and energy efficient
routing algorithms which do not use the existing SPP solution.

3 WSN Modeling

We consider wireless sensor networks with cluster-based topology which is illus-

trated in Fig. 1.
The development of energy-efficient and reliable communication protocols
is very important. Depending on the case, a WSN may consist of hundreds or
thousands of sensor nodes. Each wireless sensor node uses the protocol stack to
communicate with other nodes and the control center or the cluster head (CH).
In [13] a couple of algorithms are presented for self-organizing wireless sensor
networks. They derived an optimization problem to satisfy sensor scheduling
with the joint detection probability. The minimum energy routing problem tries
to optimize the performance of a single user (an end-to-end connection) by min-
imizing its energy consumption. The typical way to solve this problem is to use
the shortest path algorithm for wired networks. However, there is a difference
between the WSN design problem in this study and the existing WSN design
problem. The existing WSN design problem finds the shortest path in the given
graph, whereas the aim of this study is to find the shortest path between nodes
and the energy. In what follows, we present a mathematical model for the WSN
design problem that includes only distance and energy restrictions. Thus we have
developed the mathematical model for WSN with cluster topology under energy
and time resource restrictions. More specifically, we consider the maximum dis-
tance between the linking nodes and the energy of the nodes, keeping in mind
that sensors have limited available energy. A sensor transmits radio signals in
4 A. Alexandrov and V. Monov

order to connect to other nodes or to the CH. However, physically the radio
signal cannot reach long distances because of energy and frequency limitations.
Because the wireless sensors are battery powered, they also have energy limita-
tions. Therefore, each sensor is on standby in the sleep mode, which makes it
possible to minimize energy consumption. When a communication event occurs,
the sensor enters the wake-up mode and receives or transmits information. Then,
it searches for CH with the aim of sending information to the control center.
The sensor node must reach the CH. In view of the above working characteris-
tics of the network, we have developed our model of WSN under the following
assumptions:

– The location of the CH does not change during the process of collecting data
from the sensor nodes in the cluster.
– The sensor nodes GPS coordinates are fixed.
– Each sensor node has energy limitations.
– The Cluster Head (CH) has not any energy limitations.
– Every sensor node as part of the cluster has a direct connection with CH.

Fig. 1. WSN modeling with cluster topology

In Fig. 2, we schematize the notation used in mathematical modeling.

 Design of a Multi-objective Optimization Model 5

Fig. 2. Mathematical modeling notation

3.1 A Mathematical Model of WSN with Resource Restrictions

The presented WSN model is also characterized by the following requirements.

There are several sensor nodes and a cluster head node (CH) who redirect the
sensor data to other cluster head nodes or directly to the control center. The
current WSN design modeling is similar to the shortest path problem. However,
the presented model is used to find the shortest path between the sensor nodes,
CH and the control center when a cluster topology is given. A sensor consumes
energy at a fixed time period when it connects with other sensors. Therefore,
we consider only the energy used in the active mode and in the transmission
mode. The energy consumed by the sensor node in sleep mode is at a very low
level compared to the energy in active mode and in the presented model we
do not consider it. We assume that all of the sensors energy levels are initially
the same and that the transmission energy used is in direct proportion to the
distance between the sensors. In the presented model, we consider a few types of
resource limitations, those on distance, period and energy use. The model aims
to avoid severance in a single period. The objective is to minimize the sum of
the amounts of energy used.
The WSN Model for multi-objective optimization is formulated as follows:
 
M IN Ei−begin − Ei−current (1)
i∈N,i=n i∈N,i=n

 
M AX W Ti−begin − W Ti−current (2)
i∈N,i=n i∈N,i=n
6 A. Alexandrov and V. Monov

where:
Ei−begin - initial amount of energy of node i, for all i ∈ N ;
Ei−current - current energy of node for all i ∈ N ;
W Ti−begin - initial amount of work time of node i, for all i, j ∈ N ;
W Ti−current - current work time capacity of node for all i N ;
The combined objective function consisting of Eq. (1) and (2) is based of a
multi-objective optimization by looking for a minimum between the sum of the
amounts of energy used by the nodes and the sum of amounts of work time of
every sensor node (his shelf life). The restriction (3)

djn ≤ dCH , all i, n ∈ N, i = n (3)

can be replaced with the Eq. (4) if the node energy is below specified limit. In
this case, the node j will route the data through the neighbor node i instead of
direct connection to the CH.

dij + din ≥ dCH , all i, j, ∈ N (4)

Ecurrent−it ≥ 0, all i ∈ N, t ∈ T (5)

 
Ei−begin − (Ecurrent−it + Ei−active + din .Ei−send ) = 0 (6)
i∈N,i=n i∈N,i=n

 
W Tbegin − (W Tcurrent−it + Ti−active + din .Ti−send ) = 0 (7)
i∈N,i=n i∈N,i=n

Constraint (3) indicates that there must be flow from node i and into CH
node n. Constraint (3) illustrates the limit on the transmission distance and can
be replaced with (4) if the amount of current energy Ei−current is below the
limit. Constraint (5) defines the remaining energy of the nodes. The remaining
energy of a node is equal to the initial energy minus the active state (eactive
) energy and the transmission energy tsend associated with the nodes in the
cluster. The remaining shelf life of a node is equal to the initial time resource
of the node minus the active and transmission periods. The objective function
of the equations is almost linear and depend on the amount of the transmission
energy and the amount of node transmission periods as part of all the active
energy and active periods.
The notation is as follows:
i, j, k - indices of node (i, j, k = 1, 2, 3, ........, N )
N - the number of all the nodes (i, j, k ∈ N )
n - cluster head (CH) node index (n ∈ N )
t - index of periods (t ∈ T )
T - set of transmission periods (t ∈ T )
din - distance from node i to CHn ;
 Design of a Multi-objective Optimization Model 7

dij - distance from node i to node j;

dCH - the maximum communication distance between a sensor node and the
CH;
Ecurrent−it - the amount of the energy of node i, in moment t for all i ∈
N (t ∈ T );
Ei−active - the amount of energy consumed by sensor node i during the sensor
active period;
Ei−send - the amount of energy consumed by sensor node i during the trans-
mission period;
W Tcurrent−it - the amount of time work capacity of node i in moment t for
all i ∈ N (t ∈ T );
Ti−active - the amount of time when the sensor i is active;
Ti−send - the amount of time of the sensor i transmission ( Ti−send ≤
Ti−active );

3.2 Results from the Simulation Experiment

The simulation of the proposed multi-objective simulation model was realized by

the industrial OPNET 14.0 network modeler. For the simulation purposes was
used the embedded in the modeler real ZigBee based network protocol, identical
to the protocol Zigbee Smart Energy V2.0 used in real sensor devices for cluster
based networks. The simulated WSN was based on one cluster, consisting of
10 wireless network nodes in a star topology with one cluster head (CH). The
transmission period Ti−send vary between 50 ms and 150 ms. The active period
of every sensor Ti−active was fixed by 200 ms.
The Table 1 below shows the result of average cluster transmission energy
consumption Ei−send , the cluster work capacity in hours and the transmission
period Ti−send . As is shown on the diagram there is an optimum between average
node energy consumption and calculated work time capacity at an approximately
119 ms time period when the node distance is between 20–50 m and another
optimum at 111 ms when the node distance is between 50–100 m.

Table 1. Experimental data generated by OPNET 14

8 A. Alexandrov and V. Monov

Fig. 3. Optimization results based on the simulation by industrial modeler OPNET 14

4 Conclusions
This paper considered a wireless sensor network mathematical design under
resource restrictions. We developed an improved mathematical model for multi-
objective optimization of a wireless sensor network with cluster topology by
simultaneously considering different size node transmission periods and energy
limits of the sensor nodes. The proposed model can be used in real situations in
designing of battery-powered wireless sensor networks. The developed model can
also be used in other sensor networks design under various resource restrictions.
Even though the current mathematical model does not require significant compu-
tation time, much more time may be needed for large-scale WSNs. The further
research will be focused on some related simulation studies of the model and
development of other multi-objective optimization techniques including addi-
tional restrictions, related to QoS and the existing adaptive routing algorithms
as SLTP protocol implemented in real WSN.

References
1. Pottie, G.J., Kaiser, W.J.: Wireless integrated network sensors. Commun. ACM
43(5), 5158 (2000)
 Design of a Multi-objective Optimization Model 9

2. Atanasova T., Modelling of Complex Objects in Distance Learning Systems. In:

Proceedings of the First International Conference - Innovative Teaching Methodol-
ogy, 25–26 October, Tbilisi, Georgia, pp.180-190 (2014). ISBN 978-9941-9348-7-2
3. Salami, O., Bagula, A., Chan, H.A.: Framework for link reliability in inter-working
multi-hop wireless networks. Math. Comput. Modell. 53(1112), 22192228 (2011)
4. Tashev, T.D., Hristov, H.R.: Modeling of Synthesis of Information Processes with
Generalized Nets. J. Cybernetics and Inform. Technologies, Sofia, Academic Pub-
lishing House ?Prof. Marin Drinov?, no. 2, pp. 92–104 (2003)
5. Yick, J., Mukherjee, B., Ghosal, D.: Wireless sensor network survey. Comput. Netw.
52(12), 22922330 (2008)
6. Balabanov, T., Zankinski, I., Barova, M.: Strategy for Individuals distribution
by incident nodes participation in star topology of distributed evolutionary algo-
rithms. Cybern. Inf. Technol. 16(1), Sofia, 80-88 (2016). Print ISSN: 1311–9702,
Online ISSN: 1314-4081
7. Wagner, D., Wattenhofer, R.: Algorithms for Sensor and Ad-hoc Networks.
Springer-Verlag, Berlin Heidelberg, Germany (2007)
8. Eom, H.Y., Kim, C.S., Melikov, A., Fattakhova, M.: Approximation method for
QoS analysis of wireless cellular networks with impatient calls. Ind. Eng. Manage.
Syst. 9(4), 339347 (2010)
9. Bari, A., Jaekel, A., Bandyopadhyay, S.: Optimal placement and routing strategies
for resilient two-tiered sensor networks. Wirel. Commun. Mobile Comput. 9, 920937
(2009)
10. Bari, A., Wazed, S., Jaekel, A., Bandyopadhyay, S.: A genetic algorithm based
approach for energy efficient routing in two-tiered sensor networks. Ad Hoc Netw.
7(4), 665676 (2009)
11. Zhang, J., Wu, C.D., Zhang, U.Z., Ji, P.: Energy-efficient adaptive dynamic sensor
scheduling for target monitoring in wireless sensor networks. ETRI J. 33(6), 857863
(2011)
12. Domingo, M.C.: Packet size optimization for improving the energy efficiency in
body sensor networks. ETRI J. 33(3), 299309 (2011)
13. Zhao, W., Stankovic, J.A., Ramamritham, K.: A window protocol for transmission
for time-constrained messages. IEEE Trans. Comput. 39(9), 11861203 (1990)
14. Hur, S.J., Kim, J.H., Choi, J.H., Park, Y.W.: An efficient addressing scheme and
its routing algorithm for a large-scale wireless sensor network. EURASIP J. Wirel.
Commun. Netw 13 (2008)
15. Al-Karaki, J.N., Ul-Mustafa, R., Kamal, A.E.: Data aggregation and routing in
wireless sensor networks: optimal and heuristic algorithms. Comput. Netw. 53,
945960 (2009)
16. Zhua, Y.H., Wua, W.D., Pana, J., Tang, Y.P.: An energy-efficient data gathering
algorithm to prolong the lifetime of wireless sensor networks. Comput. Commun.
33(5), 639647 (2010)
17. Ahn, N., Park, S.: An optimization algorithm for minimum connected dominating
set problem in a wireless sensor network. Ind. Eng. Manage. Syst. 10(3), 221231
(2011)
18. Samadian, R., Noorhosseini, S.M.: Probabilistic support vector machine localiza-
tion in wireless sensor networks. ETRI J. 33(6), 924934 (2011)
19. Jia, J., Chen, J., Chang, G., Tan, Z.: Energy efficient coverage control in wireless
sensor networks based on multi-objective genetic algorithm. Comput. Math. Appl.
57, 1756–1766 (2009). https://doi.org/10.1016/j.camwa.2008.10.036
20. Lee, J.-H.: Ilkyeong moon modeling and optimization of energy efficient routing in
wireless sensor networks. Appl. Math. Model. 38, 2280–2289 (2014)
 Cross-Validated Sequentially Constructed
Multiple Regression with Centered Predictors

Slav Angelov1(B) and Eugenia Stoimenova2

1 Department of Informatics, New Bulgarian University, 21 Montevideo Street, Sofia, Bulgaria
slav [email protected]
2 Institute of Mathematics and Informatics, Bulgarian Academy of Sciences (BAS), Acad. G.

Bontchev Street, Bl. 8, Sofia, Bulgaria

[email protected]

Abstract. In this paper, we observe a centered version of a technique that we

have called Cross-validated sequentially constructed multiple regression. It cross-
validates the estimates of the coefficients in a multiple regression model while
merging some of the predictors into components. The Centered cross-validated
sequentially constructed multiple regression is designed to do the same, but for
a multiple regression with centered predictors. The problem is that the proposed
technique is much heavier to estimate while with centered right side. We prove a
theorem that directly links the solutions of the centered and non-centered versions
of the technique. This helps us to estimate the centered version as fast as the non-
centered one. Additionally, we give some important remarks while using cross-
validation techniques for centered data.

1 Introduction

In this paper, we observe a centered version of a technique that we call Cross-validated

sequentially constructed multiple regression (Angelov and Stoimenova 2019). It cross-
validates the estimates of the coefficients in a multiple regression model while merging
some of the predictors into components. The algorithm behind the technique handles
multicollinearity problems (see, Alin 2010, Lin 2008, Paul 2006), and is useful when
we have small datasets - it tends to reduce overfitting (see, Steyerberg 2009, Clark 2004)
because of the nature of cross-validation (Browne 2000, Picard 1984).
The centering means that all of the independent variables in the regression model are
centered by their means. There are at least two reasons to use centering of the predictors
in a regression model. First, the interpretation of the intercept term. In the non-centered
version, the value of the intercept is interpret as the output of the model if all of the
independent variables are set to 0, but some variables may not become zero. Thus, the
interpretation value of the model is increased if we center it. In this case, the value of
the intercept is the output of the model if all of the predictors are set to their mean
values. Second, the centering of the independent variables can reduce multicollinearity.
This holds when there are some added interaction terms, i.e. in the polynomial regres-
sion case where we have the predictors along with other predictors that are powers of
them. The highest-order terms are not affected, but the centering may greatly affect the
coefficients of the lower-order terms, see (Aiken and West 1991). Additionally, there
c The Author(s), under exclusive license to Springer Nature Switzerland AG 2021
I. Georgiev et al. (Eds.): BGSIAM 2018, SCI 961, pp. 10–15, 2021.
https://doi.org/10.1007/978-3-030-71616-5_2
 Cross-Validated Sequentially Constructed Multiple Regression 11

are some more specific cases where centering can be useful, e.g. the intercept has no
effect on the collinearity of the other variables if the independent variables are centered
(Belsley 1980). If you want to know more about the effects of centering, see (Dalal and
Zickar 2012, Robinson and Schumaker 2009, Jaccard 1990). All that we mention about
centering induces the need to investigate what is happening when we center the inde-
pendent variables before applying the Cross-validated sequentially constructed multiple
regression.

2 Centered Cross-Validated Sequentially Constructed Multiple

Regression
We have a multiple regression model:
Y = α0 + α1 .X1 + α2 .X2 + · · · + αk .Xk + ξ , (1)
where Y is the predicted variable, αi is the coefficient in front of predictor Xi for i =
1, . . . , k, α0 is the model intercept term, and ξ is the error term.
We interpret the value of the intercept term from Model 1 as the value of the model
if all predictors are set to zero. However, in some cases the predictors cannot take such
values which makes the interpretation of the intercept term inadequate. To solve this
issue, we can center the right side of Model 1. We define the centered model as:
Y = α0 + α1 .X1c + α2 .X2c + · · · + αk .Xkc + ξ  , (2)
where Y is the predicted variable, αi is the coefficient in front of the centered predictor
Xic = Xi − μi for i = 1, . . . , k, α0 is the model intercept term, ξ  is the error term, and μi is
the sample mean of Xi . For Model 2, the intercept ξ  can be interpret as the value of the
model when all predictors are set at their mean values. Such an interpretation is always
adequate. This makes the centered model the preferred model in a lot of problems.
In our previous paper, we have presented a technique that estimates Model 1 in a step
by step procedure that minimizes the RMSECV error - Cross-Validated Sequentially
Constructed Multiple Regression (CVSCMR). Root mean squared error after cross-
validation (RMSECV) measure:

∑ni=1 εi2
RMSECV = , (3)
n
where n is the number of the observations in the primary data set, εi is the error for the i-
th observation which is obtained from the regression model estimated over the primary
data set without the i-th observation. RMSECV consists of out of sample errors which
makes it more realistic measure for the model fit (Hawkins 2003).
Let present a brief overview of the CVSCMR technique:
1. We estimate Model 1;
2. We choose the model variable which has the worst estimate based on its absolute t-
value, let say Xi with a coefficient αi . Then we find the most correlated with it model
variable X j with a coefficient α j , and we combine them into one component by the
following way Zi, j = Xi + k ∗ X j if |αi | > |α j |, or Zi, j (k) = X j + k ∗ Xi ;
12 S. Angelov and E. Stoimenova

3. We find k while minimizing the RMSECV (k) for Model 1 with Zi, j (k) that replaces
Xi and X j (we explain in more details further in the paper);
4. Back to Step 1.
The procedure ends when we have achieved absolute t-values over a chosen threshold
for all of the derived model variables or when the model variables are reduced to a
predefined percentage from their initial number. Many other criteria are possible.
We can note that on each iteration of the algorithm the variables are reduced by 1,
and this reduction is obtained by the optimization of the RMSECV applied on Step 3.
By cross-validating some of the estimates we aim to achieve estimates that are better
estimated, and will improve the performance of the model for out-of-sample observa-
tions. Once the model is estimated, if we want, we can express the model by the initial
predictors from the obtained components.
About the procedures in Step 3, for each two chosen predictors (or already obtained
components) Xi and X j , we construct component Z(k) = Xi + k ∗ X j (assuming |αi | >
|α j |). Then we are minimizing:

∑ni=1 εi2 (k) a j + τσ j a j − τσ j
min RMSECV (k) = min , k∈( , ),
k k n ai − τσi ai + τσi
τ > 0, ai = ±τσi , (4)

where τ > 0 is a number between 1 and 6, lets say 4; n is the number of obser-
vations; α j , αi are estimated coefficients from Model 1; σi and σ j are the stan-
dard errors of αi and α j ; εi (k) = Y (i) − fi (k) is the error from the i-th observation
from the model fi (k). fi (k) is the multiple regression model with input variables
Z(k), X1 , X2 , . . . , Xi−1 , Xi+1 , . . . , X j−1 , X j+1 , . . . , Xn (we replace the predictors with the
constructed components) estimated with the whole data set except the i-th observation.
To find the optimum k, we should use a branch-and-bound algorithm (see Jiao 2013,
α
Norkin 1998, Clark and Wood 1966) to search around k0 = αij in the defined interval
(if we choose k0 we will obtain the same result as if we have used the estimates from
the multiple regression). The problem is that no matter how efficient is the implemented
algorithm on each iteration, i.e. for any chosen k, we must calculate m regressions fi (k),
m is the number of observations, in order to calculate RMSECV (k). The good thing is
that we can avoid all of the estimates of fi (k) because it is known that the out-of-
sample residuals εi , i = 1, . . . , n used in the RMSECV measure can be computed by the
following formula:
ξ̂i
εi = , i = 1, . . . , m, (5)
1 − hi
where ξ̂i is the error of the i-th observation for model Y (k) = α0 + α ∗ Z(k) + α1 ∗ X1 +
α2 ∗ X2 + . . . + αi−1 ∗ Xi−1 + αi+1 ∗ Xi+1 + . . . + α j−1 ∗ X j−1 + α j+1 ∗ X j+1 + . . . + αn Xn ,
m is the number of observations, and hi,i is the i-th element from the diagonal of the hat
matrix H = X(X t X)−1 X t , where X is the design matrix, X t is its transposed matrix.
The aim of this work is to investigate what will be the difference if we apply the
proposed algorithm for the centered model from (2). On a first glance, there will be no
difference, we just center the predictors and applied the model. The problem is that if
 Cross-Validated Sequentially Constructed Multiple Regression 13

we do this, it will cause the cross-validation procedure to be invalid one. While cross-
validating, we assume that we do not know anything about the out-of-sample obser-
vations. Thus, each time when we choose an out-of sample observation, we should
recenter the predictors using the corresponding means that are estimated without this
observation. As a consequence, we cannot use (5) to compute RMSECV (k) fast. If we
use it anyway, it will be equivalent to centering the predictors once, and ignoring any
recentering.
The centered version of the Cross-validated sequentially constructed multiple
regression follows the same algorithm as the non-centered one. The difference is in
the complication around the procedure for finding the optimal k for a chosen compo-
nent Z c (k) = Xic + k ∗ X jc . For simplicity, we assume that this is the first component that
we are obtaining from the algorithm. We search for the optimal k while minimizing:

∑ni=1 (εic (k))2 aj + τσ j aj − τσ j
min RMSECVc (k) = min , k∈( , ),
k k n ai − τσi ai + τσi
τ > 0, ai = ±τσi , (6)

where α j , αi , and their corresponding σ j , σi are from Model 2, and εic (k) = Y (i)− fic (k)
is the error from the i-th observation from the model fic (k). The functions fic (k) have
the following form:

fic (k) = αi,0 + αi ∗ Zic (k) + αi,1 ∗ (X1 − μi,1 ) + αi,2 ∗ (X2 − μi,2 ) + . . .
+ αi,i−1 ∗ (Xi−1 − μi,i−1 ) + αi,i+1 ∗ (Xi+1 − μi,i+1 ) + . . . + αi, j−1 ∗ (X j−1 − μi, j−1 )+
+ αi, j+1 ∗ (X j+1 − μi, j+1 ) + . . . + αi,n ∗ (Xn − μi,n ),

where Zic (k) is Z c (k) centered without taking into account the i-th observation, μi, j , i −
f ixed, j = 1, . . . , n are the corresponding sample means estimated without the i-th obser-
vation. Now, we note that not only we cannot use the formula from (5) for a fast com-
putation of RMSECV (k) in (6), but there are additional computations while compared
with the non-centered method if we do not use that formula.

Theorem 1. If we apply the the Cross-validated sequentially constructed multiple

regression method for the regression from Eq. 1 and for the centered regression from
Eq. 2 then the obtained sets of components from the two models will be derived by
the same functional relation ships, i.e., the difference will be that the components for
the Eq. 1 will use X1 , . . . , Xk as input arguments while the components for the centered
model will use the centered versions of X1 , . . . , Xk .

Proof. We will start with the fact that the regression model from Eq. 1 and its centered
version with centered X1 , . . . , Xk are identical in the sense that their model residuals are
the same. Moreover, their model coefficients are the same. The only difference is in the
intercept term. This fact is known from the regression analysis theory, a proof can be
seen in (Sen and Srivastava 1990, pp. 42–43).
We have the multiple regression problem from Eq. 1. We start the procedure for
obtaining components for the both models. Let us consider for simplicity that we have
chosen to combine variables X1 and X2 into a component Z(k) = X1 + k.X2 . While
14 S. Angelov and E. Stoimenova

centering the variables X1 , . . . , Xk the correlations between them will not be changed.
Thus, we will choose again X1 and X2 as building block for the first component for the
centered model case. If the centered version of X1 and X2 are X1c and X2c then from the
way we define the functional form of a component and the fact that the centered model
has the same coefficients as the non-centered one, we conclude that the component
Z c (kc ) from X1c and X2c has the same form as Z(k), i.e., Zc (k) = X1c + kc .X2c . Moreover,
the starting point k0 for the estimation of k is the same for the two components Z(k) and
Zc (k). We will define with k∗ the optimum k which will participate as the coefficient
in the formula for Z(k), and we will define with kc∗ the optimum kc in the formula for
Zc (kc ). Now, if we prove that k∗ and kc∗ coincide then Z(k∗ ) and Zc (kc∗ ) will be formed
by the same formula.
To find the optimum k for Z(k) we need to minimize RMSECV (k), to find the opti-
mum kc for Zc (kc ) we need to minimize RMSECVc (kc ). Both of this functions consist of
out-of-sample model errors. Let us observe one of these errors ε fi,k from RMSECV (k)
and its corresponding one ε fi,k c from RMSECVc (kc ). These two errors are obtained from
c
the i-th observation using models fi,k and fi,k c . The regression f has model variables
c i,k
Z(k), X3 , . . . , Xk , the regression fi,kc has model variables Zc (k), X3c , . . . , Xkc . We start the
c

optimization procedures for k and kc from one and the same point k0 , so we will observe
k and kc like one variable k. For each k and for each i fi,k c is the centered model of f
i,k
c c
because X3 , . . . , Xk are the centered versions of X3 , . . . , Xk (centered without considering
the i-th observation) and Zc (k) is the centered version of Z(k). As a consequence, for
a fixed k and i the out-of-sample i-th model error will be the same for the both mod-
els. Then for each k RMSECV (k) = RMSECVc (k). This means that the minimum of
RMSECV (k) will be reached in the same point k∗ like the minimum of RMSECVc (kc ).
Then the derived components are Z = X1 + k∗ .X2 and Zc (kc ) = X1c + k∗ .X2c , i.e., they
have the same functional form.
Following the same logic we can continue to derive components for the centered
and non-centered models and they will have the same functional form. 


Corollary 1. If we have the components for variables X1 , . . . , Xk , and we want to obtain

the components for the centered model, we will use the same components, but with the
centered X1 , . . . , Xk as building blocks. Thus, we can estimate the centered version of the
Cross-validated sequentially constructed multiple regression as fast as the non-centered
one.

3 Summary

We presented in this paper the centered version of the Cross-validated sequentially

constructed multiple regression. We stressed and showed two things. First, it is not
correct to apply the Cross-validated sequentially constructed multiple regression on
centered (or normalized) once data. This holds for any manipulation of the data that
is performed by using a statistic which is estimated from the whole dataset. Second,
it is possible to directly estimate the centered version of the technique from the non-
centered one. This saves n-times slower computation speed, where n is the number of
observations.
 Cross-Validated Sequentially Constructed Multiple Regression 15

Acknowledgements. The work of E.S. was supported by the National Science Fund of Bulgaria
under Grant KP-06-N32/8.

References
Angelov, S., Stoimenova, E.: Cross-validated sequentially constructed multiple regression. In:
Georgiev, K., Todorov, M., Georgiev, I. (eds.) Advanced Computing in Industrial Mathematics,
BGSIAM 2017. Studies in Computational Intelligence, vol. 793. Springer, Cham (2019)
Sen, A., Srivastava, M.: Regression Analysis: Theory, Methods, and Applications. Springer, New
York (1990)
Aiken, L., West, S.: Multiple Regression: Testing and Interpreting Interactions. Sage, Thousand
Oaks (1991)
Dalal, D., Zickar, M.: Some common myths about centering predictor variables in moderated
multiple regression and polynomial regression. SAGE J. 15(3), 339–362 (2012)
Jaccard, J., Wan, C., Turrisi, R.: The detection and interpretation of interaction effects between
continuous variables in multiple regression. Multivariate Behav. Res. 25(4), 467–478 (1990)
Robinson, C., Schumaker, R.: Interaction effects: centering, variance inflation factor, and inter-
pretation issues. Multiple Linear Regression Viewpoints 35(1), 6–11 (2009)
Belsley, D., Kuh, E., Welsch, R.: Regression Diagnostics: Identifying Influential Data and
Sources of Collinearity. Wiley, New York (1980)
Alin, A.: Multicollinearity. Wiley Interdisc. Rev. Comput. Stat. 2(3), 370–374 (2010)
Lin, F.: Solving multicollinearity in the process of fitting regression model using the nested esti-
mate procedure. Qual. Quant. 42(3), 417–426 (2008)
Paul, R.: Multicollinearity: Causes, Effects and Remedies. IASRI, New Delhi (2006)
Picard, R., Cook, R.: Cross-validation of regression models. J. Am. Stat. Assoc. 79(387), 575–
583 (1984)
Browne, M.: Cross-validation methods. J. Math. Psychol. 44(1), 108–132 (2000)
Hawkins, D., Basak, S., Mills, D.: Assessing model fit by cross-validation. J. Chem. Inf. Comput.
Sci. 43(2), 579–586 (2003)
Clark, T.: Can out-of-sample forecast comparisons help prevent overfitting? J. Forecast. 23(2),
115–139 (2004)
Steyerberg, E.: Overfitting and optimism in prediction models. In: Clinical Prediction Models,
pp. 83–100. Springer, New York (2009)
Lawler, E., Wood, D.: Branch-and-bound methods: a survey. Oper. Res. 14(4), 699–719 (1966)
Norkin, V., Pflug, G., Ruszczyński, A.: A branch and bound method for stochastic global opti-
mization. Math. Program. 83(1–3), 425–450 (1998)
Jiao, H., Wang, Z., Chen, Y.: Global optimization algorithm for sum of generalized polynomial
ratios problem. Appl. Math. Model. 37(1–2), 187–197 (2013)
Random documents with unrelated
content Scribd suggests to you:
West. These people belonged to the outstandingly non-conformist
type. They were sufficiently independent—contemptuous, one might
say—of established customs and institutions to be willing, with what
ignorance or awkwardness soever, to bring about changes, some of
which were sadly needed. Religiously they were apt to be come-
outers. It was largely among this class that were recruited the
Millerites, Millennialists, and original Latter Day Saints, together with
many other minor sects and factions. In politics, when all orthodox
New England was Whig, they were mainly Democratic; many,
however, backed the program of Nativism; in the person of John
Brown they exemplified the principle of direct action as applied to
slavery. The social innovator, the medical quack, and the political
demagogue found among them welcome and encouragement,
sometimes to the temporary distress of society, often to its ultimate
benefit. Not unlike the original Puritans who represented “the
dissidence of dissent and the protestantism of the protestant
religion,”[26] they constituted a dynamic social element although
wanting in the intellectual and religious training, the political morale,
and perhaps the heroism which distinguished the original planters of
Massachusetts Bay. They had the spirit of the revolutionary New
Englanders, who were described, not inaptly, as “hard, stubborn,
and indomitably intractable.” They were the backbone of Shays’s
rebellion. In many ways they illustrate the qualities which, at various
times in our later history, have served as the fulcrum of revolutionary
change.
Dwight’s foresters were merely the extreme manifestation, the
caricature, of a much larger class of heady, self sufficient,
opinionated, and troublesome persons who equally with the sober,
church going, instructed, conformist type were the product of New
England conditions. The cords of restraint were drawn so taut in the
parishes and towns, that the person who was determinedly
“different” was compelled to break them and become a kind of social
pariah in order to gain the freedom his soul craved. It was not an
accident that so large a proportion of that class went to the frontier.
They found there a less rigorous church discipline, freedom from
taxes for the support of the established church, and a more flexible
state of society in the midst of which they might hope to function. In
western Massachusetts and Connecticut, in Vermont, New
Hampshire, and Maine, they were numerous at the opening of the
National period. Soon large numbers emigrated to western New
York, to northern Pennsylvania, to Ohio, thence throughout the
West. They made up an appreciable part of the thronging Yankee
immigration which seized upon Wisconsin’s prairies and oak
openings between 1835 and 1850, and their presence has left its
impress upon our social history. Still the experiences of older
frontiers, such as western New York, had already modified the type.
When all necessary deductions have been made, however, the
church remained equally with the school a dominant note in the
Yankee’s social landscape. His “meeting house,” not infrequently in
New England a gem of ecclesiastical architecture, fulfilled his artistic
ideal; the congregation was the “household of faith” which claimed
his undeviating loyalty; the pastor was “guest and philosopher” in his
home whenever he chose to honor it with his presence. To men and
women alike, attendance upon the church services was the principal
Sabbath day duty and the chief physical and mental diversion of the
whole week. It was an old custom to linger after the morning
sermon for a social chat either in the church yard, when the weather
permitted, or else at a near-by tavern; and while the talk was
ostensibly about the sermon, gossip, bits of practical information,
and even a shy kind of love making were often interwoven, tending
to make this a genuine community social hour.
The tradition that the minister must be a man of learning was of
incalculable social importance. His advice was called for under every
conceivable circumstance of individual and community need. He
assisted about the employment of schoolmasters and was the
unofficial supervisor of the school. He enjoined upon negligent
parents the duty of sending their children, and he had an eye for the
promising boys—lads o’pairts, as the Scotch say—whom he
encouraged to prepare for professional life. He fitted boys to enter
the academy and sometimes tutored college students. In the rural
parish the minister occupied the church glebe, which made him a
farmer with the rest. He was apt to read more widely and closely in
the agricultural press, or in books on husbandry, than his neighbors,
thereby gaining the right to offer practical suggestions about many
everyday matters. Some ministers were writers for agricultural
journals. Many contributed to local newspapers items of news or
discussions of public questions in which their parishioners were
interested with themselves.
The home missionary idea was inherent in the New England system
both as respects religion and education. Older, better established
communities always felt some responsibility for the newer. Since
settlement proceeded largely by the method of planting new
townships of which the raw land was purchased by companies from
the colonial and state governments, it was possible for the larger
community to give an impetus to religion and education under the
terms of township grants. This was accomplished by reserving in
each grant three shares of the land—“one for the first settled
minister, one for the ministry forever, and one for the school.” Other
grants of raw land were made for the support of academies. Here
we have the origin of the system of land grants in aid both of the
common schools and of state universities, in the western states. The
grants for religion necessarily were discontinued after the adoption
of the national constitution.[27]
The religious unity established by the Puritans, and maintained for a
time by the simple method of rigorously excluding those holding
peculiar doctrines, gave way to considerable diversity in the late
eighteenth and early nineteenth centuries. Episcopalianism made
some progress in the older settlements, and Unitarianism created a
great upheaval, while toward the frontiers the Methodists and
Baptists flourished more and more. These several elements, by
1820, were powerful enough politically to secure the abolition of the
ancient tax for the support of the established (Congregational or
Presbyterian) church—a tax which had long caused ill feeling
between West and East, and no doubt had contributed to the growth
of dissenting churches. These frontier churches had the
characteristics of the frontier populations. Their ministers were less
learned, their morale less exacting, their religion less formal and
ritualistic, their ordinances less regularly and habitually enforced. But
there was an emotionalism which in a measure compensated for
defects of training, for looseness of habit and negligence in the
practice of religion. In a word, the camp meeting type of Christianity
prevailed widely along the frontier, and that type entered Wisconsin
Territory with the numerous Methodist and Baptist settlers from New
York and New England. As early as August, 1838, such a camp
meeting was held under Methodist leadership in the woods near
Racine; it was attended by hundreds of pioneer families drawn from
the sparsely settled neighborhoods for many miles around. Its
appointments were of the typical frontier kind, though one would
expect less boisterousness in the manifestations of emotion among
those people than seems to have accompanied similar gatherings in
the Southwest.[28]
The stated religious services in early Wisconsin, as in every frontier
region, were apt to be less frequent than in older communities.
Ministers were too few in number and neighborhoods too
impecunious to justify each locality in supporting a minister. The
circuit riding custom prevailed generally among all denominations.
One preacher traveled, on foot, six hundred miles, making the round
in six weeks. Each group of churches also had its conferences, which
were occasions for planning missionary effort, for unitedly attacking
special religious or social abuses, and for promoting constructive
community effort. The ablest speakers addressed such gatherings;
the membership of the churches concerned and others attended, in
addition to the delegates; and important religious, social, or moral
results sometimes flowed from them.
Another peculiar Yankee institution allied at once to the school and
the church, was the lyceum or local coöperative organization for
bringing lecturers to the community. The settlements in southeastern
Wisconsin had their lyceums at an early date, and many
distinguished public men from the East had occasion to visit this new
Yankeeland in the capacity of lecturer. Among them were Horace
Greeley, Bayard Taylor, and James Russell Lowell.
Reform movements, however, though usually receiving valuable aid
from churches, lyceums, mechanics’ institutes, and other permanent
organizations of men for public discussion, had a way of creating
special organizations to propagate themselves. That was true of the
temperance movement, which by the time of the Yankee
immigration into Wisconsin was under vigorous headway. Beginning,
in serious form, about 1820, the intervening years witnessed the
creation of hundreds of local temperance societies in New England
and New York, and the federation of these societies into state
societies. These central organizations stimulated the movement by
sending out lecturers, conducting a newspaper propaganda, and
issuing special publications. Some of their tracts are said to have
been scattered “like the leaves of autumn,” all over New England and
New York.
One of these tracts affected the social history of Wisconsin very
directly. It is known, traditionally, as “The Ox Discourse,” because it
was based on Exodus 21:28-29: “If an ox gore a man or woman,
that they die: then the ox shall be surely stoned, and his flesh shall
not be eaten; but the owner of the ox shall be quit. But if the ox
were wont to push with his horn in time past, and it hath been
testified to his owner, and he hath not kept him in, but that he hath
killed a man or a woman; the ox shall be stoned, and his owner also
shall be put to death.” The sermon on this text produced a great
sensation and gained many new adherents to the temperance cause.
Among these were two brothers, Samuel F. and Henry Phoenix, who
were storekeepers in a New York village and sold much whisky. They
publicly destroyed all the liquor they had on hand and became
crusaders in the temperance cause. In the spring of 1836 Colonel
Samuel F. Phoenix selected in Wisconsin a “Temperance Colony
claim,” on which he settled that summer. Then he rode to Belmont
and induced the first territorial legislature to set off from Milwaukee
County a county to be known as Walworth, in honor of the
chancellor of the state of New York, who was a noted temperance
leader. He named the village begun by him Delavan, in honor of E.
C. Delavan, pioneer temperance editor and at that time chairman of
the executive committee of the New York State Temperance Society.
Colonel Phoenix lectured on temperance, helped to organize early
temperance societies, rebuked his neighbors—especially the New
Yorkers—for employing whisky at raisings, and, before his death in
1840, had succeeded in giving a powerful impulse to the movement
in southeastern Wisconsin.
Another dramatic figure in early temperance annals was Charles M.
Goodsell, who in 1838 settled at Lake Geneva and built the first mill
operated in Walworth County. He was of Connecticut birth, and his
father owned and managed, among other properties, a whisky
distillery. Goodsell, however, when he came west from New York
State, was a most determined opponent of the traffic in intoxicants.
Soon after opening his mill a local company erected in Lake Geneva
a distillery for making corn whisky. Goodsell warned them, he says,
not to expect him to grind their grain and they installed a grinding
apparatus of their own. But, their machinery proving inadequate,
they finally sent a grist of corn to Goodsell’s mill, demanding, as
under the law they had a right to do, that it be “ground in turn.”
Goodsell refused, thereby producing a tense situation, for the
pioneer farmers looked to the distillery as a cash market for their
grain. Finally, the distillers brought suit, won a verdict, and Goodsell
appealed. But meantime, he rode to Madison, where the legislature
was sitting, and procured the adoption of an amendment to the law
regulating milling, to the effect: “Nothing in this section contained
shall be construed to compel the owners or occupiers of mills to
grind for distilling, or for sale or merchant work.” This proviso,
adopted in 1841, remained a feature of the statute for many years.
[29]

It must not be supposed that pioneer Yankee society, even in

Walworth County, was prevailingly of the temperance variety. All
testimony, both of the reformers and of others, tends to show that a
large majority was at first in the opposition. Frontier history would
indicate that excessive indulgence in whisky was apt to be more
common during the primitive phase of settlement than later, due
perhaps to the looser social and religious organization.
Wisconsin may be said to have been born to the temperance
agitation which, in a few years’ time, produced societies pledged to
total abstinence all over the southeastern part of the state and in
many other localities. In March, 1843, a legislative temperance
society was organized with a list of twenty-four signers. The house
of representatives at the time had twenty-six members, the council
thirteen, or a total of thirty-nine. So a decided majority was aligned
with the movement. Moses M. Strong was chosen president, which
was considered a triumph for the cause, and much interest was
aroused by the adherence of William S. Hamilton, who is reported to
have addressed one of the society’s meetings.[30]
The temperance agitation everywhere received a notable impetus
from the adoption in 1851 of the prohibition law by the state of
Maine. Immediately other states moved for the same objective, and
in Wisconsin a referendum vote was taken in 1853 which resulted
favorably to prohibition, though no enactment followed.[31] In that
election the southeastern counties were overwhelmingly for the
Maine law. Walworth gave 1906 votes for it and 733 against, Rock
2494-432, Racine 1456-927. Milwaukee at the same time voted
against prohibition by 4381 to 1243. This shows where was to be
found the powerful opposition to legislation of this nature, which was
destined to increase rather than diminish with the strengthening of
the German element already very numerous.
From the time of the Maine law agitation the communities
dominated by Yankees were generally found arrayed in favor of any
proposal for limiting or suppressing the liquor traffic, although, as we
shall see in later articles, no large proportion of their voters ever
joined the Prohibition party. They did not succeed in abolishing
drunkenness, though it became very unfashionable to indulge
heavily in spirituous liquors and the proportion of total abstainers
among the younger generation steadily increased. Yankees furnished
a very small per cent of those who gained their livelihood through
occupations connected directly with intoxicating liquors, except as
such traffic was carried on incidentally as a feature of the drug
business. The disfavor with which saloon keeping, brewing, and
distilling have long been regarded among that class of the
population is explained by the fervor and thoroughness of the early
temperance campaigns.
Because of their attitude on the liquor question, on Sunday laws,
and other matters pertaining to the regulation of conduct, the
Yankees have always been looked upon by other social strains as
straight-laced and gloomy. In this judgment men have been
influenced more than they are aware by the traditions of Puritanism
which it was supposed the Yankees inherited. They recalled the story
of how Bradford stopped Christmas revelers and sent them to work;
they pictured Puritan children as forbidden to laugh and talk on the
Sabbath day; and some may have heard the story of how
Washington, while president, was once stopped by a Connecticut
tithing man who must be informed why His Excellency fared forth on
the Lord’s Day instead of resting at his inn or attending public
worship.[32]
Two remarks may be made on this point. First, while Puritanism
unquestionably had a somber discipline, there was not lacking even
among Puritans the play instinct which persisted in cropping out
despite all efforts of the authorities at repression. Second, the
nineteenth century Yankees register a wide departure from early
Puritanism in their social proclivities, and the difference was
particularly marked in the West. Even church services were modified
to fit the needs of the less resolute souls. Music became an
important feature and it was adapted more or less to special
occasions.[33] Sunday Blue Laws were gradually relaxed, though
never abandoned in principle. Well-to-do city people allowed
themselves vacation trips, visits to watering places, and to scenic
wonders like Niagara Falls.[34] In town and country alike dancing
became an amusement of almost universal vogue, though protested
by some religionists, and rural neighborhoods found bowling such a
fascinating game for men and boys that the almanac maker thought
well to caution his readers against over-indulgence therein.[35] Ball
playing, picnicing, sleighing, coasting, skating were among the
outdoor sports much indulged in by Yankees, while family and
neighborhood visiting, the quilting bee, donation parties, church
socials, and the like furnished indoor recreation. The circus and the
“cattle show” were events in the western Yankeeland equal in social
significance to Artillery Day in Boston.
Thus, while it is true that Yankees were a sober people, of
prevailingly serious mien and purpose, they were not averse to the
relaxations of play and recreation. The question whether or not the
Yankees were fun loving cannot be answered by yes or no. If we
mean by fun the rollicking joviality characteristic of irresponsible,
carefree folk, the answer is no. Many Yankees found their best fun in
work or business. To the David Harum type, which was fairly
numerous, a horse trade was more fun than a picnic. Some Boston
merchants were so immersed in their business that, though very
pious, they nevertheless spent Sunday afternoon going over their
books and writing business letters.[36] Being serious minded, they
tended to make their chief concern an obsession, and could hardly
be happy away from it. But the majority were quite as ready to
amuse themselves out of working hours, as are the Italians or other
social stocks that have a reputation for fun and frolic.
The Yankees also found intellectual enjoyment in cultivating
quickness of retort, in giving utterance to clever if homely
aphorisms, and in a kind of whimsical humor. These traits emerge in
their vernacular literature like “Major Jack Downing’s” Thirty Years
out of the Senate, and especially Lowell’s Biglow Papers. “The
squire’ll have a parson in his barn a preachin’ to his cattle one o’
these days, see if he don’t,” said one of “Tim Bunker’s” shiftless
neighbors by way of summarizing the squire’s over-niceness in
caring for his Jersey cows. “Ez big ez wat hogs dream on when
they’re most too fat to snore”; “that man is mean enough to steal
acorns from a blind hog”; “the coppers ain’t all tails”; “pop’lar as a
hen with one chicken”; “quicker’n greased lightnin’”; “a hen’s time
ain’t much”; “handy as a pocket in a shirt”; “he’s a whole team and
the dog under the wagon”; “so thievish they had to take in their
stone walls at night”; “so black that charcoal made a chalk mark on
him”; “painted so like marble that it sank in water”—the above are
all Yankeeisms of approved lineage and illustrate a characteristic
type of Yankee humor. The example below is of a rarer sort. “Pretty
heavy thunder you have here,” said the English Captain Basil Hall to
a lounger in front of a Massachusetts tavern. “Waal, we do,” came
the drawling reply, “considerin’ the number of the inhabitants.”
About the time that Yankees began to emigrate to Wisconsin a
talented French writer, Michel Chevalier, gave the world a brilliant
and on the whole favorable characterization of them. “The Yankee,”
he says, “is reserved, cautious, distrustful; he is thoughtful and
pensive, but equable; his manners are without grace, modest but
dignified, cold, and often unprepossessing; he is narrow in his ideas,
but practical, and possessing the idea of the proper, he never rises
to the grand. He has nothing chivalric about him and yet he is
adventurous, and he loves a roving life. His imagination is active and
original, producing, however, not poetry but drollery. The Yankee is
the laborious ant; he is industrious and sober and, on the sterile soil
of New England, niggardly; transplanted to the promised land in the
west he continues moderate in his habits, but less inclined to count
the cents. In New England he has a large share of prudence, but
once thrown into the midst of the treasures of the west he becomes
a speculator, a gambler even, although he has a great horror of
cards, dice, and all games of chance and even of skill except the
innocent game of bowls.” Chevalier also says: “The fusion of the
European with the Yankee takes place but slowly, even on the new
soil of the west; for the Yankee is not a man of promiscuous society;
he believes that Adam’s oldest son was a Yankee.”
The Yankee was not more boastful than other types of Americans,
though his talent for exaggerative description was marked. Yet he
had a pronounced national obsession and was uncompromising in
his patriotism: “This land o’ourn, I tell ye’s got to be a better country
than man ever see,” was put into a Yankee’s mouth by one of their
own spokesmen and represents the Yankee type of mild jingoism. It
is full cousin to that other sentiment which also this writer assigns to
him:
Resolved, that other nations all, if set longside of us,
For vartoo, larnin, chiverlry, aint noways wuth a cuss.[37]
These are but cruder expressions of ideas dating from the
Revolutionary War, and of which Timothy Dwight, who was not a
poet by predestination, gave us in verse a noble example:
Columbia, Columbia, to glory arise,
The queen of the world, and child of the skies!
Thy genius commands thee; with rapture behold,
While ages on ages thy splendors unfold.
Thy reign is the last, and the noblest of time,
Most fruitful thy soil, most inviting thy clime;
Let the crimes of the east ne’er encrimson thy name,
Be freedom, and science, and virtue, thy fame.
It need not be supposed that all Yankees who came to Wisconsin or
other western states were familiar with these glowing lines. But it is
almost certain that, in the common schools of Yankeedom, most of
them had thrilled to the matchless cadences of Webster’s reply to
Hayne. What more was needed, by way of literary support, to a
pride of country which, if a trifle ungenerous to others, was based
on facts all had experienced.
THE YANKEE AND THE TEUTON IN WISCONSIN
Joseph Schafer

IV. SOME SOCIAL TRAITS OF

TEUTONS
The year 1832, celebrated in Wisconsin history as the time when the
lead miners and other pioneers destroyed the power of the Rock
River Indians, was remembered by later-coming German immigrants
for a very different reason. It was toward the end of March in that
year, the place Trier (Treves), the ancient capital of the western
“Cæsars,” a city which is still rich in the massive ruins of its Roman
foretime. As the story goes, the boys of one form in the old
Gymnasium were being entertained at the house of a professor,
where, boy-like, they were playing indoor games accompanied with
much laughter and general hilarity. Suddenly one of their younger
classmates rushed breathless into the room, exclaiming: “Goethe is
dead!”[38] During the balance of the evening, the less serious of the
youngsters having returned to their interrupted play, this boy
engaged with his instructors in eager discussion of Goethe’s life and
writings.
The youth in question was Karl Marx, whose later history exhibits a
wide divergence from the exclusively literary career prophesied by
his boyhood scholastic interests. The classmate who is authority for
this incident continued in Marx’s company the Gymnasium studies;
he then performed his one year minimum of military service, and
having secured some business experience sailed away as an
immigrant to the new world, settling on a Wisconsin farm. In the
course of a long life he often reverted to the story of Goethe, whose
works, as well as those of Schiller and Lessing, made a part of his
home library. These great names never failed to kindle his pride in
the intellectual achievements of the German people, whose
governments at the time of his emigration in 1841 seemed to him a
compound of despotism and inefficiency.[39]
Doubtless there were Germans of the immigration to Wisconsin who
knew not Goethe, or if in a hazy way they did know who he was,
had no intellectual right to judge his merits. But the more intelligent
were sure to possess some knowledge of the writings of their
greatest poet and of lesser men who still were great in the world’s
estimation. Hence it was that Germans who at that period went to
the new world, while acknowledging by their flight the political,
economic, and social obstacles to a successful life in Prussia,
Bavaria, Baden, Westphalia, or Luxemburg, were always able to
maintain a self-respecting attitude when confronted with the
pretensions of those Americans who were unsympathetic, jingoistic,
or boastful. German immigrants might grant much to superior
cleverness, to the stupendous achievements of a liberty loving race,
domiciled in a peaceful continent and dowered with free lands and
boundless opportunity; but they remembered that William Tell and
Faust and The Laocoön were written by Germans.
Though many immigrants were far from being literary, they
doubtless possessed, on the average, a knowledge of German
masterpieces fully equivalent to the knowledge which Americans
possessed of the English Classics. For education was looking up, and
while most of the immigrants from German states, like those from
other European countries, were of the peasant class, which was
usually the most backward, still by 1840 nearly all were sure to have
enjoyed some systematic schooling. At an earlier period this might
have been otherwise. The condition of limited serfdom, removed but
a generation earlier, operated powerfully to neutralize such
benevolent plans for universal instruction as kings and ministers
proclaimed. For the peasants were directly subordinate to the local
lords, who often felt “that an ignorant labor supply was less likely to
seek to better its condition by demands upon them....”[40] The great
national reform movement which came to fruition after the close of
the Napoleonic wars swept away many of the disabilities of the
common people, and developed in Prussia and other states a system
of universal education as the surest means of national upbuilding.
The excellencies of the Prussian school system prior to 1840 became
the theme of flattering reports on the part of educators in many
lands. The celebrated philosopher Victor Cousin made it the basis for
his plan of educational reform in France; the Scotch, English, and
Irish discussed it; Horace Mann proclaimed it to the school
authorities of Massachusetts, and Calvin E. Stowe recommended it
to the legislature of Ohio. That system may not have possessed all
of the virtues which the ordinances quoted by Cousin imply.[41] Yet it
had the one excellence to which educationally all others are
subsidiary—a well-trained teaching force. Indeed, if there is anything
which seems miraculous in the swift and thoroughgoing
transformation of school conditions in Prussia during the first forty
years of the nineteenth century, it is explained by the provision
which the state made for normal schools and the supply, through
their agency, of teachers enough to man all the schools. “In the
lowest school in the smallest and obscurest village,” says Horace
Mann, “or for the poorest class in overcrowded cities; in the schools
connected with pauper establishments, with houses of correction or
with prisons—in all these there was a teacher of mature age, of
simple, unaffected and decorous manners.” Mann also made it clear
that every such teacher was possessed of adequate scholarship and
special training for the work of the schoolroom.[42] Such a statement
could not be made at that time about Massachusetts, where popular
education was already two hundred years old, nor could it be made
with equal confidence of other German countries, though several of
these approximated the Prussian standard and most of them were
earnestly promoting education along the same lines and by the use
of similar means.
We must therefore regard the generation of the German exodus
from which Wisconsin profited so largely in the later 1840’s and the
1850’s, as almost universally literate and usually well grounded in
the rudiments of an education. The intelligent, reading, writing, and
slow but careful figuring German peasant immigrants constituted the
best testimonial to the efficacy of German systems of instruction for
the common people. The Gymnasia, the real Schule, the universities,
sent forth representatives of the highest German culture to honor
the learned professions, the literary, philosophical, and scientific
circles of America.
On the basis of formal school instruction alone, the historian of early
Wisconsin would be compelled to assign first place in social fitness
to the immigrants from Germany. Neither the Irish, the English, nor
even the Yankee pioneers on the average had enjoyed as thorough a
training as had Prussians, Saxons, Hessians, or Badeners. Yet, school
training is never all there is of education, and it may constitute but a
small portion of it. No one questions that the social character of
Prussian and other German peasants was far higher in 1840 than it
had been in 1800, and this was due to a variety of causes, of which
schooling was only one. In part it was due to the abolition of
serfdom, in part to the reorganization of municipal life; also, largely
to the religious agitation of the period, to the movements for political
reform, and especially to the widespread, momentous, and gripping
spirit of nationalism.
Nevertheless, despite their superb educational equipment plus other
incentives, the Prussians still seemed to intelligent American
observers in a very retarded social condition. Horace Mann, who
wrote most enthusiastically of their schools and was sympathetic
toward the Germans in every respect, in a passage of almost classic
force and beauty written in 1843, tells us why education in Prussia
accomplished for the people so much less than one might expect.
For one thing, he says, the pupils left school too early—at the age of
fourteen, which was their time for beginning regular and heavy
work. Then, too, books for further self-instruction were lacking.
There was in Prussia nothing analogous to the Massachusetts district
school libraries. “But,” he continues, “the most potent cause of
Prussian backwardness and incompetency is this—when the children
come out from the school they have little use either for the faculties
that have been developed, or for the knowledge that has been
acquired. Their resources have not been brought into demand; their
powers are not roused or strengthened by exercise. Our common
phrases, ‘the active duties of life’; ‘the responsibilities of citizenship’;
‘the stage, the career, of action’; ‘the obligations to posterity’;—
would be strange sounding words in the Prussian ear.... Now,
although there is a sleeping ocean in the bosom of every child that is
born into the world, yet if no freshening, life-giving breeze ever
sweeps across its surface, why should it not repose in dark
stagnation forever.” The bill of particulars with which the great
educator clinches his indictment of the Prussian system, while it
aims to describe accurately only the then existing condition in
Prussia, might be equally applicable to almost any other absolutist,
paternalistic state. All responsibility for the people’s welfare was
assumed by the monarch, who in turn was actively aided by a
hierarchy of officials in state and church, in the central government
and the local administrative areas.
Of this officialdom, particularly in its military and civil aspects, the
nobility was not merely the corner stone but the essential part of the
structure. The church, loyal to its traditions, was much more
democratic, men of every class being found in each of its official
grades. The newly developed educational system gave to the
common man another significant opportunity, since teaching
candidates were drawn in large numbers from the middle and lower
classes, and were given at public expense the training necessary to
fit them for permanent positions in the various types of schools. On
the whole, however, life beyond the school, which among Americans
of that day commonly yielded the major part of education, was in
Prussia far less fruitful. For, the American, whose formal schooling
had been limited, was sure to multiply its efficacy many times
through the intensely original character of his activities. In these he
was apt to employ everything he had learned, and constantly to
learn more for the sake of applying the new knowledge to
challenging situations.
The contrast between the average Prussian’s life and the average
American’s life was sharp and decisive. The boy leaving school at
fourteen in Frederick William’s country was thrust at once into a
routine of severe labor, controlled by others. Either he might be on a
farm, where his duties were fixed by custom and minutely directed
by parent or employer; or he might be apprenticed to a trade which
would give him seven years under an exacting master. Assuming
that he remained in his native region, his career thenceforth would
be determined with the minimum of personal effort. The American
boy whose schooling stopped at an early age might go west and
start a new farm home in a new environment, with every incentive
toward employing his best powers to win unusual success; he might
go to the city and engage in some business; attend school to
prepare for a profession; or settle down on the ancestral acres under
social and economic conditions which called for almost continuous
readjustments, and kept his mind on the stretch to bring these
about.
The governmental arrangements in America were inherently
educational; in Prussia they were the reverse, save when, with
revolutionary fury, the people rose to seek their destruction or
reform. In Prussia, says Horace Mann, “the subject has no officers to
choose, no inquiry into the character or eligibility of candidates to
make, no vote to give. He has no laws to enact or abolish. He has no
questions about peace or war, finance, taxes, tariffs, post office, or
internal improvements to decide or discuss. He is not asked where a
road shall be laid, or how a bridge shall be built, although in the one
case he has to perform the labor and in the other to supply the
materials.... The tax gatherer tells him how much he is to pay, the
ecclesiastical authority plans a church which he must build; and his
spiritual guide, who has been set over him by another, prepares a
creed and a confession of faith all ready for his signature. He is
directed alike how he must obey his King and worship his God.”
The schools of Prussia inculcated religion and morality as sedulously
as they taught geography, singing, and writing, the methods used
being highly praised by American pedagogical experts. This universal
insistence on the ethical content of life could not fail to produce
results more or less in harmony with the aims of great ethical
philosophers, like Kant of Königsberg, a teacher of the learned
whose “categorical imperative,” popularized in that epoch, has not
yet gone into the philosophical discard. The average German
immigrants of the 1840’s knew little of Kant or the Kantian school of
ethics. But of honesty, truthfulness, and fidelity to the plighted word
they knew much, because those were practical virtues with which in
school if not at home all were indoctrinated. Thrift and industry were
additional but fundamental virtues which were widely diffused. It is
hard for an empty sack to stand upright. The reason why in America
a German’s note was more often worth face value than that of some
other classes was because the German usually labored unceasingly
and saved what he earned, thus enabling him to meet his obligation.
[43]

They were not all saints, these Germans, and in the matter of
personal morality the Prussians particularly seem in those days to
have deserved much of the criticism directed against them.[44]
However, it is not necessary to regard even the Prussians as more
lax than most other continentals, and their character is always
explainable as a vulgarized aping of the low if gilded immoralities of
court and aristocracy. Matters of this sort do not lend themselves
readily to statistical inquiry. But it can hardly be doubted that in
France, Prussia, Austria, or any other country of continental Europe
the private morals of the common people were better on the whole
than those of the upper classes. In America, where immigrants from
those countries came into contact with a self-governing people of
simple habits and prevailingly high ideals of personal conduct,
though with numerous individual divergences from the type, sharp
attention was bound to be directed to this feature in the character of
foreigners, and the Germans attracted their full share of suspicion
and disfavor from the stricter sort of Americans.
Such suspicions were heightened by certain social customs of the
Germans to which Americans reacted adversely. Sunday
amusements were all but universal among them. Travelers in
Germany dwell upon the gaiety observed in the villages, or in the
city parks and the beer gardens, the distinctive costumes of different
localities lending color and interest to the scenes. Music was
cultivated in every German community; all Germans could sing and a
large proportion could perform on musical instruments. One was “as
certain to see a violin as a blackboard in every schoolroom.”[45]
Wherever Germans gathered together—and Sunday, since it was the
weekly holiday, was their day for assembling—there was singing and
dancing, usually accompanied by the drinking of beer or wine to
stimulate hilarity. This drinking was not necessarily excessive,
because most Germans were moderate in their appetites for alcohol,
some were unable to spend much, and all were economical
(sparsam). The dances differed from those favored in this country,
being mainly “round dances,” and the standards of decorum in the
relations of the sexes were different also. No wonder that, when
German families settled in groups near our own people, Yankee
fathers and mothers often shook their heads doubtfully in
contemplating the influence upon their children of these unfamiliar
social customs.
It is probable that the vigor with which among this resilient people
amusements were carried on had a definite relation to the intensity,
monotony, and sordidness of the labor from which they were a
recoil. At all events, with more leisure on week days and an
opportunity to do his work under pleasanter conditions, the German
readily adapted himself to a type of relaxation which was less
boisterous and more genteel. His work and his living being what
they were, it is doubtful if anything better in the form of
amusements could have been expected of him. Travelers from
England and America, on their visits to Germany, were impressed
with the wholesomeness of the Sunday picnics, the rambles through
the forests, the frolics on the village greens and in the parks
adjacent to the towns and cities.[46]
With all his sociability, joviality, and occasional levity, the German
was not devoid of an element of austerity. This was one secret of his
ability to achieve. Whatever the work might be, he settled himself to
its performance with a grim determination expressive of century-
long training. The mechanic, from his apprentice years, was
habituated to long hours of unremitting but improving toil. The
farmer (bauer) was a traditional daylight-saver and a night-worker
besides, such excessive labor being compulsory under the system of
serfdom, when the peasant’s time was levied upon to a very large
extent by the lord. The German schools inculcated similar habits of
relentless application to the work in hand, and even the government
bureaus, under rigorous task-masters like old Friedrich Wilhelm and
his son Frederick the Great, enforced compliance with the ideal of a
patient, steady “grind” which not inaptly typified the German in the
eyes of other peoples. The German often performed less work in the
time consumed than an alert Yankee would have performed in a
shorter day; his tools and implements were generally awkward and
inefficacious; even in scholarship he not infrequently took the long
way around to reach his goal—but he usually reached it because he
had no notion of turning back or of stopping at a halfway point on
his job. Persistent rather than brilliant, more industrious than
inventive, the German toiled on, content if he always had something
to show for his labor. The contrast, in that generation, between the
German at work and the German at play is the contrast between a
man governed by an intense purpose to accomplish a given task,
whether interesting or not, and the same man intent on
accomplishing nothing with every physical, intellectual, and
emotional evidence of enjoying the process. Some men carry into
their play the morale which governs them in their work; others
import into their work the spirit of their play. In the case of the mid-
nineteenth century German the two aspects of his existence, work
and play, differed in spirit quite as much as in content.
The Germans had their Puritan sects, like the Moravians and other
pietists, whose attitude was distinctly other-worldly, to whom play
was a sedate if not a solemn activity. Such people disapproved of
dancing and beer drinking Germans quite as heartily as of profane
whiskey drinking and quarrelsome Americans or Irish. Individuals
and colonies of the pietistic classes passed into the emigrations, and
thus Wisconsin’s German population contained most of the elements
to be found at the same time in the German states. This illustrates
one difficulty in generalizing about social characteristics; there are so
many exceptions to be noted that the generalization loses much of
its validity.
Craftsmanship was a prevalent accomplishment among the Germans
of the early emigration. Every shipload of emigrants of which we
have a social analysis had a large proportion of craftsmen, who were
either established members of the city and village industrial class, or
else belonged to the peasantry and had learned a craft in order to
improve their status. Trades were learned exclusively under the
apprenticeship system, the candidate usually living in the master’s
home and giving service at the master’s will. When he reached the
journeyman stage he was privileged to find work for himself, a quest
which though usually fruitful in educational results often proved
disappointing from a monetary point of view. In those cases the
journeyman was peculiarly open to the temptation to emigrate.
Arrived in this country, the chances of finding employment in the line
of his training varied. Sometimes they were excellent, at other times
poor, depending mainly upon the craft represented. Carpenters were
in great demand, as were also blacksmiths, wheelwrights,
millwrights, masons, bricklayers, plasterers, and in general all
representatives of the building trades and of trades ministering to
farmers. Others were in occasional demand. But, if a dyer, or a
slater, or a cabinet maker, or a silversmith, or a tile maker, or a
weaver, or a wood carver happened to find himself in America
without a market for his peculiar skill, he always had the resource of
taking land and commencing as a farmer. Many craftsmen, indeed,
came with the set purpose of doing that immediately upon their
arrival; others contemplated a farming career after a period devoted
to their specialty. In some or all of these ways Germans trained as
craftsmen came to be widely distributed over the farming areas of
Wisconsin as well as among the cities, towns, and villages.
The possession of special skill in any line, like the possession of
special scientific knowledge, raises a man in social estimation, and
every trained worker properly regards himself with satisfaction as
being not quite “as other men are.” In addition to the social training
which came to him as an incident of his apprenticeship and
journeyman’s experience, the German craftsman often was able to
challenge the respect and admiration of his American neighbors by
making articles of cunning workmanship which to them seemed
wonderful because they did not understand the processes involved.
Agriculture being regarded as an unskilled occupation, the artisan
farmer also was very apt to lord it over the peasant farmer of his
own nationality. Craftsmanship, in a word, established a kind of rank
among Germans in this country because it was a recognized means
of personal and social progress at home.
Statistics are impossible to procure, but the testimony of men and
women familiar with early conditions in Wisconsin proves that the
German population of the state in early days varied quite as widely
in social characteristics as did the American population, though
America had no distinctive peasant class. Accordingly, although in
the beginnings of American contacts with their Prussian or
Westphalian neighbors these were lumped together indiscriminately
as “Dutchmen,” differences soon began to emerge. In the course of
a few years a class of “fine old Germans” was recognized in almost
every community to supplement the well-known type of “fine old
Yankee gentlemen.” These select Germans were very apt to be men
who had been trained as craftsmen, or men who had enjoyed the
advanced scientific or literary instruction afforded in the higher
schools and the universities of the homeland. In the cities, especially
Milwaukee, were many Germans who had been prominent in
business lines as well as in the professions.
The question has sometimes arisen why so many of the second-
generation Germans appear inferior in social character to their
immigrant parents. A hint of the reason is found in what has just
been said. Whatever elements of superiority were shown by the
immigrant artisan-farmers or the highly educated Germans, the
social advantages accruing therefrom were personal, and in a
slightly developed western society could not be handed on to the
next generation. In the cities it frequently was possible for men of
high ideals and fine social status to provide equivalent opportunities
for their children. But not so on frontier farms. There it was a rare
case when an education or training like that received by the father in
the old country could be supplied. Accordingly, the sons of the most
intelligent, dignified, and worthy German farmer, if they became
farmers in succession, might perhaps turn out mere farmers, with
none of the graces or exceptional social virtues of the parents, and
little except the memory of a parent’s high respectability to
distinguish them from the farmer sons of the clumsiest peasant.
However, this is but half the story. If the superior Germans reared
families incapable of remaining on their own social plane, other
types of Germans, who in their own persons counted for less,
frequently had the happiness to see their children advance to a
position perceptibly higher than their own. Natural gifts, industry, the
social opportunities which yield to the key of economic success
availed much. Sometimes the presence of a good school, a wise and
helpful pastor or some other worthy friend gave the necessary
impulse. The process, in fact, does not differ essentially from that
which, throughout American pioneer history, has enabled the
deserving to press forward and permitted the weak, indolent, or
vicious to fall behind in the social competition. It is impossible to say
how many German families made a step, or several steps, upward,
and how many others slipped back. The delinquents may perhaps
exceed the meritorious in number, but probably not, and the
impression that the children of German immigrants shame their
parents is almost certainly an illusion which would be likely to
disappear if the facts were fully known.
The social institutions of Wisconsin, based on the earlier Yankee and
southwestern immigrations, were profoundly influenced by the
German immigration of the late forties and the fifties of last century.
Milwaukee, the center of German influence (the Deutsche Athen),
became a city in which the German language was spoken and read
by many English speaking persons, in order to facilitate
communication and trade with the numerically dominant German
element. The Germans maintained advanced schools for instruction
in both English and German; their parochial schools were conducted
mainly in German; the immigrants themselves felt no compulsion to
learn English, and their children, in many cases, however well
educated, spoke the language of the country with very imperfect
accent.
The universal respect in which the German language was held, and
the extent to which it was affected by others than Germans,
provided an admirable social soil for the development of German
music and the cultivation of German literature. Hardly had the
immigrants established themselves when, in 1847, they founded at
Milwaukee their first singing society, which was followed three years
later by the famous and far-reaching Musikverein. A German theater
followed promptly, and became a permanent feature of Milwaukee’s
intellectual life.[47] The Turnverein fostered in America Father Jahn’s
conception of athletics, while restaurants and beer gardens gave an
old world, continental atmosphere to public recreation. Holidays
assumed a German aspect. The Christ Child displaced St. Nicholas
not alone in Milwaukee, but in scores of towns, villages, and
hamlets, and innumerable farm homes scattered over Wisconsin.
The joyous German Weinacht made way easily against the more
somber Puritan Christmas, which, however, had already brightened a
good deal in its progress from the seventeenth century to the
nineteenth century.
In general, Germans did not insist with extreme pertinacity upon the
retention of their own social customs, and wherever people of that
nationality were intermingled with a larger number of Americans, the
process by which they assimilated American habits of living,
American social usages, and even ways of acting, speaking, and
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com