Business

Analytics
Module 3
 Linear
Optimization
Optimization via Mathematical Programming

Mathematical Programming

• A family of tools designed to help solve managerial problems in which the decision maker
must allocate scarce resources among competing activities to optimize a measurable goal

Optimal solution: The best possible solution to a modeled problem

Linear programming (LP): A mathematical model for the optimal solution

of resource allocation problems. All the relationships are linear.
Application Case 9.5

Spreadsheet Model
Helps Assign Medical
Residents
• Company
• Problem description
• Proposed solution
• Results
LP Problem Characteristics
Limited quantity of economic resources

Resources are used in the production of products or services

Two or more ways (solutions, programs) to use the resources

Each activity (product or service) yields a return in terms of the goal

Allocation is usually restricted by constraints

Linear Programming Steps
Decision variables
Objective function

Identify the … Objective function coefficients

Constraints
• Capacities / Demands / …

LINDO: Write mathematical formulation

Represent the model EXCEL: Input data into specific cells in Excel

Run the model and observe the results

 Product-mix problems (how many of each
product to produce for max profit)

Transportation (minimize cost of shipments)

Common Assignment (best matching of objects)

Optimization
Models Investment (maximizing rate of return)

Network optimization models for planning

and scheduling

Replacement (capital budgeting), …

Multiple Goals, Sensitivity
Analysis, What-If Analysis,
and Goal Seeking
Multiple Goals
• Multiple Goals
• Simple-goal vs. multiple goals
• Vast majority of managerial problems
has multiple goals (objectives) to
achieve
• Attaining simultaneous goals
• Methods of handling multiple goals
• Utility theory
• Goal programming
• Expression of goals as constraints,
using LP
• A points system
Multiple Goals
• Certain difficulties may arise when
analyzing multiple goals
• Difficult to obtain a single organizational
goal
• The importance of goals change over
time
• Goals and sub-goals are viewed
differently
• Goals change in response to other
changes
• Dynamics of groups of decision makers
• Assessing the importance (priorities)
Sensitivity Analysis
• Sensitivity analysis
• It is the process of assessing the
impact of change in inputs on outputs
• Helps to …
• eliminate (or reduce) variables
• revise models to eliminate too-
large sensitivities
• adding details about sensitive
variables or scenarios
• obtain better estimates of sensitive
variables
• alter a real-world system to reduce
sensitivities
• Can be automatic or trial and error
What-If Analysis, and Goal Seeking

What-if analysis Goal seeking

Assesses solutions based on changes in Backwards approach, starts with the goal and
variables or assumptions (scenario analysis) determines values of inputs needed
What if we change our capacity at the milling Example is break-even point determination
station by 40% [what would be the impact] • In-order to break even (profit = 0), how many
products do we have to sell each month
Big Data Analytics
 Big Data - Definition and Concepts

Big Data means different things Traditionally, “Big Data” = Where does the Big Data come
to people with different massive volumes of data from?
backgrounds and interests
E.g., volume of data at CERN, NASA, Google, Everywhere! Web logs, RFID, GPS systems,
… sensor networks, social networks, Internet-
based text documents, Internet search
indexes, detail call records, astronomy,
atmospheric science, biology, genomics,
nuclear physics, biochemical experiments,
medical records, scientific research, military
surveillance, multimedia archives, …
 The Data Size Is Getting
Big, Bigger…
• Hadron Collider - 1 PB/sec
• Boeing jet - 20 TB/hr
• Facebook - 500 TB/day
• YouTube – 1 TB/4 min
• The proposed Square Kilometer
Array telescope (the world’s
proposed biggest telescope) – 1
EB/day
 Big Data is a misnomer!

Big Data - Big Data is more than just

Characteristics “big”

Volume
Variety

The Vs that define Velocity

Veracity
Big Data Variability
Value
…
 UNIFIED DATA ARCHITECTURE
System Conceptual View

A High-Level ERP MOVE MANAGE ACCESS

Marketing
Marketing
Executives

Conceptual SCM
DATA
PLATFORM Applications
Operational

Architecture for
Systems
CRM
INTEGRATED
DATA WAREHOUSE Customers

Big Data
Business
Partners
Images Intelligence

Solutions
Frontline
Audio Workers
and Video Data
Mining

Business
Machine
Logs DISCOVERY PLATFORM Analysts
Math
and Stats
Data
Text Scientists
EVENT
PROCESSING Languages
Web and Engineers
Social

BIG DATA ANALYTIC

SOURCES TOOLS & APPS USERS
 Big Data by itself, regardless of the
size, type, or speed, is worthless
Fundamentals of
Big Data
Big Data + “big” analytics = value
Analytics

With the value proposition, Big Data

also brought about big challenges
• Effectively and efficiently capturing, storing, and
analyzing Big Data
• New breed of technologies needed (developed
or purchased or hired or outsourced …)
 You can’t process the amount of data that you want to because
of the limitations of your current platform.

Big Data You can’t include new/contemporary data sources (e.g., social
media, RFID, Sensory, Web, GPS, textual data) because it does
Considerations not comply with the data storage schema.

You need to (or want to) integrate data as quickly as possible to

be current on your analysis.

You want to work with a schema-on-demand data storage

paradigm because of the variety of data types.

The data is arriving so fast at your organization’s doorstep that

your analytics platform cannot handle it.
 A Clear
business need

Personnel with Strong,

advanced committed
Critical analytical skills sponsorship

Success Keys to Success

Factors for with Big Data
Analytics
Alignment

Big Data The right

analytics tools
between the
business and IT

Analytics
strategy

A fact-based
A strong data
decision-making
infrastructure
culture
 • A clear business need (alignment with the vision and the
strategy).
• Business investments ought to be made for the good of
Following are the business, not for the sake of mere technology
advancements. Therefore, the main driver for Big Data
the most analytics should be the needs of the business at any
level-strategic, tactical, and operations.
critical • Strong, committed sponsorship (executive champion).

success • It is a well-known fact that if you don't have strong,

committed executive sponsorship, it is difficult (if not

factors for Big impossible) to succeed.

• Alignment between the business and IT strategy.
Data analytics • It is essential to make sure that the analytics work is
always Supporting the business strategy, and not other
way around. Analytics should play the enabling role in
successful execution of the business strategy.
 • To create a fact-based decision-making culture,
senior management needs to:
• Recognize that some people can't or won't adjust
Following are • Be a vocal supporter

the most • Stress that outdated methods must be

discontinued

critical • Ask to see what analytics went into decisions

• Link incentives and compensation to desired
success behaviors
• A strong data infrastructure.
factors for Big • Data warehouses have provided the data

Data analytics infrastructure for analytics. This infrastructure is

changing and being enhanced in the Big Data era
with new technologies. Success requires
marrying the old with the new for a holistic
infrastructure that works synergistically.
Enablers of Big Data Analytics
• Storing and processing the complete data set in
In-memory analytics RAM

• Placing analytic procedures close to where data is

In-database analytics stored

Grid computing & • Use of many machines and processors in parallel

MPP (MPP - massively parallel processing)

• Combining hardware, software, and storage in a

Appliances single unit for performance and scalability
Challenges of Big Data Analytics
The ability to capture, store, and process the huge
Data volume
volume of data in a timely manner

Data integration The ability to combine data quickly/cost effectively

The ability to process the data quickly, as it is

Processing capabilities
captured (i.e., stream analytics)

Data governance (… security, privacy, access)

Skill availability (… data scientist)

Solution cost (ROI)

Application Case 13.3
eBay’s Multi Data-Center Deployment
eBay’s
Big Data
Solution

Questions for Discussion

1. Why did eBay need a Big Data solution?
2. What were the challenges, the proposed solution, and the obtained
results?
BUSINESS PERFORMANCE MANAGEMENT
(BPM)

Business Performance Management (BPM) is…

• A real-time system that alerts managers to potential opportunities, impending

problems and threats, and then empowers them to react through models and
collaboration.

Also called corporate performance management (CPM by

Gartner Group), enterprise performance management (EPM by
Oracle), strategic enterprise management (SEM by SAP)
BUSINESS PERFORMANCE MANAGEMENT
(BPM)
BPM refers to the business processes, methodologies, metrics, and
technologies used by enterprises to measure, monitor, and manage
business performance.

BPM encompasses three key components

• A set of integrated, closed-loop management and analytic processes, supported by

technology …
• Tools for businesses to define strategic goals and then measure/manage performance
against them
• Methods and tools for monitoring key performance indicators (KPIs), linked to organizational
strategy
A CLOSED-LOOP
PROCESS TO
OPTIMIZE BUSINESS
PERFORMANCE
Decision Analysis
 Decision analysis (DA) is a form of decision-making
that involves identifying and assessing all aspects of
a decision, and taking actions based on the decision
that produces the most favorable outcome.

The goal of decision analysis is to ensure that

decisions are made with all the relevant information

Decision and options available.

Analysis
For example, a corporation may use it to make
million-dollar investment decisions, or an individual
can use it to decide on their retirement savings.

As a form of decision-making, the fundamentals of

decision analysis can be used to solve a multitude of
problems, from complex business issues to simple
everyday problems.
 • A relatively new approach to
supporting decision making
• a.k.a. Decision Automation
Automated Systems (DAS)
• Often a rule-based system that
Decision provides a solution in a functional
area
Systems • “If only 70 percent of the seats
on a flight from LA to NY are sold
3 days prior to departure, offer a
discount of x to nonbusiness
travelers”
• Applies to repetitive/structured
decisions
Automated
Decision-
Making
Framework
 Decision analysis allows corporations to evaluate and model the
potential outcomes of various decisions to determine the correct
course of action.

How
To be effective, the business needs to understand multiple aspects of a
problem to result in a well-informed decision.

Decision
The analysis entails understanding various goals, outcomes, and
uncertainties involved, including the use of probabilities to measure
the expected outcome of various decisions.

Analysis
One of the most important aspects involves framing the problem in a
way that allows for further analysis.

Framing is typically the first part of decision analysis, and it involves

Works
creating a framework to evaluate the problem from multiple
perspectives.

They can include opportunity statements, action items, and measures

of success.

Once the framework is established, a model can be developed to

evaluate the favorability of various outcomes. Examples of models are
decision trees and influence diagrams.
 After creating a framework to evaluate a problem, models

Decision Trees are typically used to evaluate the outcomes of various

decisions. Models are visual representations of expected
outcomes, and they are used to illustrate decisions in
comparison to other alternatives.

By modeling the various expected outcomes and their

probabilities, businesses can then select the decision that
produces a favorable outcome.

One of the most common models involved in decision

analysis is decision trees, which are tree-shaped models
with “branches” that represent potential outcomes.

Decision trees are used because they are simple to

understand and provide valuable insight into a problem by
providing the outcomes, alternatives, and probabilities of
various decisions. This makes it easy to evaluate which
decision results in the most favorable outcome.
Decision Tree
Decision Tree
Real-World
• Let’s assume that a clothing store is opening a second
location and wants to decide whether to open in San
Francisco or New York. Opening a location in either city
Example will involve different capital expenditures and
demonstrate different rates of success.
• Before constructing a decision tree, we will need to
gather relevant data:
 After gathering data, we can
Real-World Example construct the decision tree based
on each decision:
 For each decision, the decision tree also includes
numerical data to calculate the expected value.
Squares represent decisions, and circles represent
outcomes. The lines branching from squares are
possible choices, while the lines branching from
circles are expected outcomes.

Decision The model also includes the costs associated with

opening each location. To open in San Francisco, the

Trees store will need to invest $2 million, while a New York

location will require an investment of $5 million.

The expected payoff amounts represent the potential

revenue if the store succeeds, or the potential loss if
the store fails.
Expected Value (EV)

• After a model is constructed, it is important to find the expected value (EV) to evaluate which
decision results in the most favorable outcome.
• Recall that the decision trees provide all the possible outcomes in comparison to the alternatives.
By calculating the expected value, we can observe the average outcomes of all decisions and then
make an informed decision.
• To calculate the expected value, we require the probability of each outcome and the resulting
value. The formula for the expected value is as follows:
• EV = (Probability A * Expected Payoff A) + (Probability B * Expected Payoff A)
• The formula above assumes that a business decision has two outcomes – success or failure. Each
outcome can be represented by Probability A or B. The Expected Payoff refers to the gain or loss
expected with each outcome.
• If there are multiple decisions to be made, a business will calculate the expected value for each
decision to determine which is most favorable.
Expected Value (EV)
• To evaluate which decision is more favorable, we will
calculate the expected value for each decision.
• EV = (Probability A * Expected Payoff A) +
(Probability B * Expected Payoff A)
• EV (San Francisco) = (0.4 * $15,000,000) + (0.6 * -
$4,000,000) = $3,600,000
• EV (New York) = (0.3 * $30,000,000) + (0.7 * -
$10,000,000) = $2,000,000
• Then, we must deduct the initial capital expenditure
to find the net gain/loss:
• San Francisco: $3,600,000 – $2,000,000 = $1,600,000
• New York: $2,000,000 – $5,000,000 = -$3,000,000
Application Case 11.1

Giant Food Stores Prices the Entire Store

• Company background
• Problem description
• Proposed solution
• Results
Modeling and Decision Making -
Under Certainty, Uncertainty,
and Risk
• Certainty
• Assume complete knowledge
• All potential outcomes are known
• May yield optimal solution
• Uncertainty
• Several outcomes for each decision
• Probability of each outcome is unknown
• Knowledge would lead to less uncertainty
• Risk analysis (probabilistic decision making)
• Probability of each of several outcomes
occurring
• Level of uncertainty => Risk (expected value)
 Modeling and Decision Making -
Under Certainty, Uncertainty, and Risk
The Zones of Decision Making