AFRICA BUSINESS SCHOOL – UM6P

Supply Chain Analytics-Machine Learning Course

Global Supply Chain Management Program - GSCM
11th and 12th November 2023

2
 Introduction

• Introduction to advanced analytics

• Introduction to Supply Chain Analytics
• Introduction to R and R Studio

3
 Introduction to advanced analytics: Agenda
• Decision Making
• Business Analytics Defined
• A Categorization of Analytical Methods and Models
– Descriptive Analytics
– Predictive Analytics
– Prescriptive Analytics
• Big Data
• Business Analytics in Practice

4
Introduction to advanced analytics
Chapter 1

5
 Introduction (Slide 1 of 3)

• Three developments spurred recent explosive growth in

the use of analytical methods in business applications:
• First development:
– Technological advances—scanner technology, data collection
through e-commerce, Internet social networks, and data
generated from personal electronic devices—produce incredible
amounts of data for businesses.
– Businesses want to use these data to:
▪ improve the efficiency and profitability of their operations,
▪ better understand their customers,
▪ price their products more effectively,
▪ and gain a competitive advantage.
6
Introduction (Slide 2 of 3)

• Three developments spurred recent explosive growth

in the use of analytical methods in business
applications (cont.):
• Second development:
– Ongoing research has resulted in numerous methodological
developments, including:
▪ Advances in computational approaches to effectively handle and
explore massive amounts of data.
▪ Faster algorithms for optimization and simulation.
▪ More effective approaches for visualizing data.

7
Introduction (Slide 3 of 3)

• Three developments spurred recent explosive

growth in the use of analytical methods in
business applications (cont.):
• Third development:
–The methodological developments were paired with an
explosion in computing power and storage
capability.
–Better computing hardware, parallel computing, and
cloud computing have enabled businesses to solve big
problems faster and more accurately than ever before.

8
 Decision Making (Slide 1 of 3)
• Managers’ responsibility:
To make strategic, tactical, or operational decisions.
– Strategic decisions:
▪ Involve higher-level issues concerned with the overall direction of the
organization.
▪ Define the organization’s overall goals and aspirations for the future.
– Tactical decisions:
▪ Concern how the organization should achieve the goals and
objectives set by its strategy.
▪ Are usually the responsibility of midlevel management.
– Operational decisions:
▪ Affect how the firm is run from day to day.
▪ Are the domain of operations managers, who are the closest to the
customer.
9
 Decision Making (Slide 2 of 3)

Decision making can be defined as the following

process:
1.Identify and define the problem.
2.Determine the criteria that will be used to evaluate
alternative solutions.
3.Determine the set of alternative solutions.
4.Evaluate the alternatives.
5.Choose an alternative.

10
 Decision Making (Slide 3 of 3)

Common approaches to making decisions include:

–Tradition.
–Intuition.
–Rules of thumb.
–Using the relevant data available.

11
Business Analytics Defined (Slide 1 of 2)

Business analytics:
–Scientific process of transforming data into
insight for making better decisions.
–Used for data-driven or fact-based decision
making, which is often seen as more
objective than other alternatives for decision
making.

12
 Business Analytics Defined (Slide 2 of 2)

Tools of business analytics can aid decision

making by:
–Creating insights from data.
–Improving our ability to more accurately forecast for
planning.
–Helping us quantify risk.
–Yielding better alternatives through analysis and
optimization.

13
A Categorization of Analytical Methods
and Models

• Descriptive Analytics
• Predictive Analytics
• Prescriptive Analytics

14
A Categorization of Analytical Methods
and Models (Slide 1 of 8)
Descriptive Analytics:
• Descriptive analytics: Encompasses the set of
techniques that describes what has happened in the past;
examples include:
– Data queries.
– Reports.
– Descriptive statistics.
– Data visualization (including data dashboards).
– Data-mining techniques.
– Basic what-if spreadsheet models.
• Data query: A request for information with certain
characteristics from a database.

15
A Categorization of Analytical Methods
and Models (Slide 2 of 8)
Descriptive Analytics (cont.):
• Data dashboards: Collections of tables, charts, maps,
and summary statistics that are updated as new data
become available.
• Uses of dashboards:
– To help management monitor specific aspects of the company’s
performance related to their decision-making responsibilities.
– For corporate-level managers, daily data dashboards might
summarize sales by region, current inventory levels, and other
company-wide metrics.
– Front-line managers may view dashboards that contain metrics
related to staffing levels, local inventory levels, and short-term sales
forecasts.
16
A Categorization of Analytical Methods
and Models (Slide 3 of 8)
Descriptive Analytics (cont.):
• Data mining: The use of analytical techniques for better
understanding patterns and relationships that exist in
large data sets.
• Examples of data-mining techniques include:
– Cluster analysis.
– Associations analysis
– …

17
A Categorization of Analytical Methods
and Models (Slide 4 of 8)
Predictive Analytics:
• Predictive analytics: Consists of techniques that use
models constructed from past data to predict the future or
ascertain the impact of one variable on another.
• Survey data and past purchase behavior may be used to
help predict the market share of a new product.

18
A Categorization of Analytical Methods
and Models (Slide 5 of 8)
Predictive Analytics (cont.):
• Techniques used in Predictive Analytics include:
–Linear regression.
–Time series analysis.
–Data mining is used to find patterns or relationships
among elements of the data in a large database; often used
in predictive analytics.
–Simulation involves the use of probability and statistics to
construct a computer model to study the impact of
uncertainty on a decision.

19
A Categorization of Analytical Methods
and Models (Slide 6 of 8)
Prescriptive Analytics:
• Prescriptive Analytics: Indicates a best course
of action to take:
–Provide a forecast or prediction, but do not provide a
decision.
–A forecast or prediction, when combined with a rule,
becomes a prescriptive model.
–Prescriptive models that rely on a rule or set of rules
are often referred to as rule-based models.

20
A Categorization of Analytical Methods
and Models (Slide 7 of 8)
Prescriptive Analytics (cont.):
Model Field Purpose
Portfolio models Finance Use historical investment return data to determine the mix of
investments that yield the highest expected return while
controlling or limiting exposure to risk.
Supply network Operations Provide the cost-minimizing plant and distribution center locations
design models subject to meeting the customer service requirements.
Price-markdown models Retailing Use historical data to yield revenue-maximizing discount levels and
the timing of discount offers when goods have not sold as planned.

• Optimization models: Models that give the best decision subject to

constraints of the situation.

21
A Categorization of Analytical Methods
and Models (Slide 8 of 8)
Prescriptive Analytics (cont.):
–Simulation optimization: Combines the use of
probability and statistics to model uncertainty with
optimization techniques to find good decisions in highly
complex and highly uncertain settings.
–Decision analysis:Used to develop an optimal
strategy when a decision maker is faced with several
decision alternatives and an uncertain set of future
events.

22
Big Data (Slide 1 of 7)

• Big data: Any set of data that is too large or too

complex to be handled by standard data-
processing techniques and typical desktop
software.
• IBM describes the phenomenon of big data
through the four Vs (as shown in Figure 1.1):
– Volume.
– Velocity.
– Variety.
– Veracity.
23
Big Data (Slide 2 of 7)

Figure 1.1: The 4 Vs of Big Data

Source: IBM

24
 Big Data (Slide 3 of 7)

Volume:
• Because data are collected electronically, we are able to
collect more of it.
• To be useful, these data must be stored, and this storage
has led to vast quantities of data.

Velocity:
• Real-time capture and analysis of data present unique
challenges both in how data are stored and the speed
with which those data can be analyzed for decision
making.
25
Big Data (Slide 4 of 7)

Variety:
• More complicated types of data are now available and are
proving to be of great value to businesses.
– Text data are collected by monitoring what is being said about a
company’s products or services on social media platforms.
– Audio data are collected from service calls.
– Video data are collected by in-store video cameras and used to
analyze shopping behavior.

• Analyzing information generated by these nontraditional

sources is more complicated in part because of the
processing required to transform the data into a numerical
form that can be analyzed.
26
Big Data (Slide 5 of 7)

Veracity:
• Veracity has to do with how much uncertainty is in the
data.
• Inconsistencies in units of measure and the lack of
reliability of responses in terms of bias also increase the
complexity of the data.

27
Big Data (Slide 6 of 7)
• Represents opportunities.
• Presents challenges in terms of data storage and
processing, security, and available analytical
talent.
• The four Vs have led to new technologies for big
data processing through distributed storage and
processing on clusters of computers.

28
Big Data (Slide 7 of 7)
• Data security, the protection of stored data from
destructive forces or unauthorized users, is of critical
importance to companies.
• The complexities of the 4 Vs have increased the demand
for analysts, but a shortage of qualified analysts has
made hiring more challenging.
• More companies are searching for data scientists, who
know how to process and analyze massive amounts of
data.
• The Internet of Things (IoT) is the technology that
allows data, collected from sensors in all types of
machines, to be sent over the Internet to repositories
where it can be stored and analyzed.
29
Business Analytics in Practice
• Financial Analytics

• Human Resource (HR) Analytics

• Marketing Analytics

• Health Care Analytics

• Supply-Chain Analytics

• Analytics for Government and Nonprofits

• Sports Analytics

• Web Analytics

30
Business Analytics in Practice (Slide 1 of 11)

Figure 1.2: The Spectrum of Business Analytics

Source: Adapted from SAS

31
Introduction to Supply Chain
Analytics

40
 Introduction to Supply Chain Analytics: Agenda

• What is Supply Chain Analytics?

• Four common areas of supply chain analytics

• Trends in Supply Chain Analytics

• How to Use Analytics in Supply Chain

• Why is Supply Chain Analytics Important

• Supply Chain Analytics Examples

41
What is Supply Chain Analytics?
• Any supply chain is a consolidated entity with many moving parts involved,
generating and demanding data to ensure effective operations at every leg of
transportation.
– Therefore, a failure within one link in a supply chain affects the entire chain, translating
into operational disruptions and unmet client expectations.
– As a result, enterprises look for practical solutions to improve the visibility of supply
chain processes and bring added value to essential services.

• Supply chain analytics implies a group of methods companies apply to

extract data from interconnected systems to get summarized data on current
and forecasted supply chain performance.

• The information can be derived from production, fleet management,

shipping, warehouse and inventory system, fulfillment software, and more.

42
 The process of managing a supply chain
• It encompasses a wide range of
interconnected activities that can be
categorized into several groups:
– Planning covers demand forecasting
and resource planning.
– Procurement is an act involved in
selecting vendors, negotiating the
terms of collaboration, and
purchasing supplies needed for your
business.
– Manufacturing concerns production
and capacity management.
– Inventory management aims at
controlling the optimal stock balance,
sales, and warehousing procedures.
– Logistics management addresses
order fulfillment and all delivery-
related activities.

43
 44

Four common areas of supply chain

analytics

44
Four common areas of supply chain
analytics
• Descriptive supply chain analytics employs dashboards and
reports to interpret what has happened. It implies using different
statistical methods to look through, summarize and structure data
about supply chain operations.
• This type of data processing is proper when answering questions
like:
– How have stock levels altered over the last month?
– What is the return on invested funds?
– But it is almost useless in forecasting: it doesn’t tell us why the
change happened or predict the customers’ expectations.

45
Four common areas of supply chain
analytics
• Diagnostic supply chain data analytics regulates cause-and-effect
relationships. It answers the following questions:
– Why are cargoes being delayed or lost?
– Why is our enterprise not fulfilling the same number of inventory
turns as a competitor?

• First, you strive to understand what's going on in your logistics

business,
• and then you are carrying out some diagnostic and analytics.

• Finally, it assists in identifying factors that affect your revenue.

46
Four common areas of supply chain
analytics

• Predictive supply chain analytics goes deeper into the

data and allows forecasting supply chain performance
under certain conditions.
– For instance, "How will a pandemic lockdown impact
the availability and price of raw materials or finished
goods?" It helps model business scenarios and
detects possible risks.
– So, this type of data processing is a more specific
niche, allowing business leaders not only to analyze
information but also to make predictions driven by the
data collected.

47
Four common areas of supply chain
analytics

• Prescriptive supply chain analytics is the peak of

technological advancement. With the help of AI (Artificial
intelligence) and ML (Machine Learning) powered
systems, businesses can predict outcomes and prescribe
the best course of actions employing optimization or
incorporated decision logic.
– This can be helpful to adjust decisions about when to
launch a product,
– whether or not to build a factory
– or define the best shipping strategy for each retail
location.

48
Trends in Supply Chain Analytics

• Internet of Things (IoT)

• Cloud Computing
• Artificial Intelligence

49
Trends in Supply Chain Analytics

• Internet of Things (IoT)

– Adding IoT functionality to the supply chain allows logistics
companies to collect different kinds of data from various supply
chain stages.
– It enables monitoring individual units of items and finished goods
inventory in the most effective way.
– Using sensors, supply chain leaders can automatically gather
data such as pressure, temperature, GPS location, etc., and
autonomously upload this information to a cloud server via the
internet.

50
Trends in Supply Chain Analytics

• Cloud Computing
– The visibility issues in supply chain operations
increase the need for cloud technology that enables
the immediate accessing and sharing of data across
platforms and projects.
– In addition, maintaining visibility allows transportation
firms to save time, free up resources, avoid issues
related to inventory shortages and order backlogs, as
well as enhance a company’s overall customer
experience.

51
Trends in Supply Chain Analytics

• Artificial Intelligence
– As big data grows, there is a need to leverage
applicable analytics to make more informed
decisions. This is where AI comes into play.
– Along with automating things like warehouse and
inventory management, AI systems that comprise
predictive analytics and machine learning algorithms
are becoming critical for smooth improvements
across planning and decision maintenance systems.

52
How to Use Analytics in Supply Chain?
• How Supply Chain Analytics Operates:
– Supply chain analytics consolidate data across multiple apps,
infrastructure, third-party services, and advanced technologies
– The process of making supply chain analytics usually starts with data
scientists who comprehend a specific aspect of the business, such
as the factors that involve cash flow, inventory, waste, and service
levels.
– Next, these specialists search for potential relations between various
data components to create a robust predictive model that streamlines
the output of the supply chain.
– After, the analytics models that met the test are deployed into
production by data engineers.
– These models are polished over time by correlating the performance
of data analysis models in manufacture with the business value they
deliver.

53
 55

Why is Supply Chain Analytics Important?

55
Why is Supply Chain Analytics Important?

• Increased Planning Accuracy

– For a logistics company, predicting consumer demand is
critical since it affects all the other supply chain processes
and is the basis for planning procurement, production
capacities, logistics, sales, etc.
– By scrutinizing consumer data, supply chain analytics will
assist businesses in better future demand prediction.
▪ For example, it will allow enterprises to determine what
goods can be minimized when they become less profitable
or decide what products will be in demand after the initial
order.

56
Why is Supply Chain Analytics Important?

• Inventory Optimization
– Analytics allows companies to strike the right stock balance to
keep costs as low as possible without stock-outs.
– For example,
▪ the system might initiate an alert for SKUs that are running
out based on the average lead time for that supplier.
▪ Sales trends can also assist the operations team in
determining which items require additional warehouse
space and which could be stored in lower numbers or
phased out.

57
Why is Supply Chain Analytics Important?

• Improved Risk Management

– Analytics solutions allow organizations to identify and predict
their potential/future risks, enhance planning, streamline
inventory management, and meet their clients’ high
expectations.
– For instance, an analytics platform could point to the risks by
noting that a specific carrier has repeatedly been late,
delivering cargoes late over the last month. It can not only
capture this pattern but signify the likelihood of further delays.

58
Why is Supply Chain Analytics Important?
• Margin Improvements
– Most logistics services providers focus on increasing revenue,
neglecting increasing profitability. Analytics can help engage
fundamental changes and challenge how things have been
performed in the past.
– Analytics systems allow identifying and analyzing costs
across the entire chain, from network costs to last mile
delivery.
▪ An essential product in this area is supply chain network
optimization, where this network is analyzed in terms of
cost to serve and service levels.
▪ Then, you can access and compare scenarios covering all
predictable market dynamics to determine the best
available option.

59
 60

Supply Chain Analytics Examples:

60
Supply Chain Analytics Examples:
• Analytics in Planning:
– Predictive analytics and machine learning techniques can be applied to support the planning
process. Thus, companies will be able to understand the current performance, past trends,
existing risks, and possible future scenarios.
– Forecasting consumer demand is figuring who will purchase your product/service, for what
price, where, and in what quantities.
– There are two basic approaches to demand forecasting:
▪ Conventional statistical methods make predictions rely on historical data and presume the continuation of
existing trends. However, the forecast's accuracy is unreliable given the fast-changing environment of most
markets.
▪ Machine learning techniques scrutinize the big data from multiple sources, recognize hidden patterns and
unobvious correlations between variables, and generate complex models that can be retrained to adjust to
evolving conditions automatically.
– Besides planning, performing deep research of consumer tastes and sales trends would
also help supply chains cut expenses on reverse logistics.
– Distribution management also relies on market research as selecting the right distribution
channels is essential for better customer reach.
– Predictive pricing strategies: By forecasting demand for the product, it’s possible to adjust
prices dynamically to what the market can bear,

61
Supply Chain Analytics Examples:
• Analytics in Procurement:
– After we understand how much needs to be purchased, SC decision makers
have to define where or from whom to buy. For this purpose, they should
scrutinize the suppliers and their performance.
– Supplier assessment is essential if SC decision makers want to ensure the
reliability and credibility of their future partners. The information that
procurement management software creates will allow to scrutinize potential
suppliers’ performance by comparing different variables, such as KPIs,
prices, compliance, etc.
– Supplier performance review means scrutinizing current suppliers' metrics
throughout your partnership lifecycle. It's vital to track vendors'
accountability, customer services, and possible additional fees. SC decision
makers can remove the poor-performing suppliers by assessing the timing
of delivery and quality of delivered goods. Moreover, supply chain analytics
techniques enable the prediction of possible supply disruptions (e.g.,
delivery delays) so that SC decision actors can take preventive actions.

62
Supply Chain Analytics Examples:
• Analytics in Manufacturing:
– Monitoring production performance: Analytics and BI tools can merge and visualize
all the critical data (e.g., machine performance, shift performance, quality, and scrap
rates) that would allow you to track your production process more effectively.
– Cost control: Cost analysis techniques such as provide a clear understanding of the
product’s cost structure and disclose opportunities for lessening your expenses.
– Optimizing maintenance: Data mining techniques allow finding correlations between
apparently disconnected factors. For instance, root cause analysis would help
determine the cause and effect between machine breakdown and low-quality spare
parts or fuel. More advanced analytics allow performing preventive maintenance
based on information from IoT devices and multiple sensors connected to equipment.
– Quality Management: Scrutinizing both quantitative and qualitative information from
manufacturing operations and consumer feedback, you get valuable insights into
enhancing your product's quality and design to meet clients’ expectations. Besides,
you can reveal the reasons behind item defects, even if they arise years later.

63
Supply Chain Analytics Examples:
• Analytics in Inventory Management:
• Predictive analytics allows organizations to determine optimal
inventory levels to satisfy demand while minimizing stock.
– Stock management: The statistics claim that US retailers lost $82 billion in 2021 due to
being out-of shelf items. The supply chain analytics monitor sales trends and inventory left,
submit suggestions on replenishments, recognize top-selling and underperforming goods,
and so on. That enables preventing costly overstocking and stockouts.
– Optimizing warehousing processes: Analytics software can scrutinize the goods flow to
provide recommendations on the best possible resources allocation and the most effective
route to pick them.
– Measuring channel performance: Analytics and BI tools allow generating different reports
to visualize your dealers' performance, products, sales channels, etc., and assist in
improving inventory control and further decision-making.
– Pricing optimization: implies adopting your prices under different internal and external
factors, such as market conditions, competitor actions, input costs, etc. Tracking sales
information along with these variables would allow you to select the optimal pricing strategy
and increase revenue.

64
Supply Chain Analytics Examples:
• Analytics in Logistics:
• Shipping and transport costs often account for a significant percentage of the final
product price. Using predictive analytics, it’s possible to determine optimal shipping
frequency and quantity to meet demand while minimizing costs.
– Fuel management. By scrutinizing driver behavior using telematics devices, SC
decision makers are able to reduce fuel costs (which can be a significant saving since
fuel costs take one-fourth of total operational expenses). smart sensors can monitor
fuel consumption, tire pressure, driving style and vehicle condition.
– Predictive-route-planning and route optimization. Using weather and traffic data
allows determining more fuel-cost-efficient routes between each stop. It helps logistics
companies maximize completed deliveries, considering many criteria such as driver
schedules, available hours, total stops, etc.
– Shipment tracking and vehicle maintenance. Gathering information from sensors,
cameras, and other IoT devices allows cargo monitoring (so that DC stakeholders can
be aware of the location and condition of the goods) and helps carry out
preventive/predictive maintenance.
– Managing returns. Recognizing the reasons that cause returns would allow to find
ways to eliminate their number. What are the reasons? For instance, the product
description doesn’t correspond to authenticity, or there is some item defect. Whatever
the reason is, it should be removed.

65
66