Scribd
Upload
380 views14 pages

Neural Networks for Effective Data Mining

This document discusses research on using neural networks for effective data mining. It begins with an introduction about how data is growing rapidly and data mining can uncover hidden patterns in large datasets. It then defines data mining and discusses how artificial neural networks can help bridge the gap between human and computer abilities for tasks like generalization. The document outlines different types of neural network models and describes the general data mining process and how neural networks can be incorporated. It concludes that neural networks are well-suited for solving data mining problems due to their robustness, self-organization, parallel processing and fault tolerance.

Uploaded by

Vandana Malkani
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
380 views14 pages

Neural Networks for Effective Data Mining

This document discusses research on using neural networks for effective data mining. It begins with an introduction about how data is growing rapidly and data mining can uncover hidden patterns in large datasets. It then defines data mining and discusses how artificial neural networks can help bridge the gap between human and computer abilities for tasks like generalization. The document outlines different types of neural network models and describes the general data mining process and how neural networks can be incorporated. It concludes that neural networks are well-suited for solving data mining problems due to their robustness, self-organization, parallel processing and fault tolerance.

Uploaded by

Vandana Malkani
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd

RESEARCH ON A NEW EFFECTIVE

DATA MINING BASED ON NEURAL


NETWORKS

Guided by Presented by
[Link] Shenoy Vandana
[Link].K
INTRODUCTION
Data is growing at phenomenal rate
We are drowning in data, starving for knowledge!
Users expect more sophisticated information
How?

UNCOVER HIDDEN INFORMATION


DATA MINING
DATAMINING DEFINITION
Tool to extract (non-trivial, implicit, previously unknown and
potentially useful) information or patterns from data in large
databases.

A hot buzzword for a class of techniques that find patterns in data.

Similar terms
 Exploratory Data Analysis
 Data driven discovery
 Deductive learning
Artificial Intelligence for Datamining

Neural Network is an abstract computer model of


human brain
Useful for datamining
People are good at generalizing from experience
Computers excel at following explicit instructions over
and over
Neural Network bridge this gap by modeling, on a
computer the neural behavior of human brains
BIOLOGICAL BACKGROUND

i/p
i/p i/p
A Neural Network generally 0
1
n
maps a set of inputs to a set of
outputs
Number of inputs and outputs is Neural
variable Network

The network itself composed of


arbitrary number of nodes with
o/p o/p o/p
arbitrary topology 0 1 n
NEURAL
NEURALNETWORK
NETWORKMODELS
MODELS
1). FEED FORWARD NETWORKS:
 - Regards the perception BP model as representatives.
 - Mainly used in areas such as prediction and pattern
recognition.
2).FEEDBACK NETWORKS:
 - Regards Hopfield model as representatives.
 - Mainly used for associative memory and
optimization calculation.
Continued……

3).SELF ORGANIZATION NETWORK:


 - Regards adaptive Resonance Theory model and
Kohonen model.
 - Used for cluster analysis.
4). RANDOM NEURAL NETWORK:
 - Developed recently.
 - Regards biological neural mathematical model.
 - Used in image processing and the combinational
optimization
GENERAL DATA MINING PROCESS
DATA MINING PROCESS BASED ON
NEURAL NETWORK
STRUCTURE OF FUZZY NEURAL
NETWORK
First layer is input layer
Second compute affiliation
relation of the input data
Third is data
preprocessing layer
Fourth is rules option
Fifth is compute the result
and output it
EFFECTIVE COMBINATION OF
KNOWLEDGE PROCESSING AND
NEURAL COMPUTATION
Evaluating:
Whether the datamining implement algorithm can adapt
complex circumstances
Whether the model is easy to understand by special user
Whether the model has good interactive characteristic
CONCLUSION

At present, data mining is a new and important area of


research, and neural network itself is very suitable for solving
the problems of data mining because its characteristics of good
robustness, self-organizing adaptive, parallel processing,
distributed storage and high degree of fault tolerance.
The combination of data mining method and neural network
model can greatly improve the efficiency of data mining
methods, and it has been widely used. It also will receive more
and more attention.
 
Bibliography
[Link]
[Link]
H Lu, R Setiono, H Liu. Effective Data Mining Using
Neural Network. IEEE Transactions on Knowledge and
Data Engineering
Xiuai Zhang, The Research on Data Transform Baseon
Data Minning. Science & Technology Information[J].
THANK YOU
THANK YOU

You might also like