OPTIMIZATION MODELS:

PRESCRIPTIVE ANALYTICS
 Learning Objective:
At the end of the lesson, the student should be able to:
• Discuss the purpose of prescriptive analytics;
• Recall how to solve linear programming models;
• Solve integer optimization models; and
• Apply prescriptive analytics to some business solutions.
 Prescriptive Analytics
Previously, we focused on descriptive and predictive
analytics to extract value from data. Descriptive
techniques were used to explore the data, and
predictive methods were implemented to classify or
predict new records.

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Prescriptive Analytics
Now, we focus on prescriptive analytics where
analytical models are developed, and the results are
analyzed to improve decision-making.
Prescriptive analytics use optimization algorithms
and simulation to quantify the effect of different
possible actions of a decision-maker to make a more
informed decision.

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Prescriptive Analytics (Jaggia et al, 2021)

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Optimization with Linear Programming
Recall from your Introduction to Management Science
Course that Optimization is a group of quantitative techniques
where an objective function is maximized or minimized for
obtaining optimal solutions to complex business problems.
For example, the objective may be to maximize profits or
minimize costs subject to resource constraints, such as fixed
amount of available materials, limited financial resources.

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Optimization and Linear Programming

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Performing Linear Programming
The first step in LP is the formulation of a problem into a
series of mathematical expressions.
Four Essential Components of the mathematical
representation of most LP problems:
1. OBJECTIVE FUNCTION. In any optimization problem, this
refers to a single objective with a mathematical
representation. The OF is assumed to be linear. The
general types are maximization function and minimization
function.

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Steps in Performing Linear Programming
2. DECISION VARIABLES. Decision variables refer to the
different choices or alternatives from which a decision
maker has to choose, in order to maximize or minimize the
value of the objective function.
3. CONSTRAINTS. These represent the limited resources
available to the decision-maker within which he/she needs
to operate.

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Steps in Performing Linear Programming
5. PARAMETERS (or input parameters) are numerical values
associated with the objective function, decision variables,
and constraints.

For example, a delivery truck driver can only work up to 14

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Some Important Concepts

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 1:

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 1: (continued)

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 1: Blending Problem (continued)
Objective: To minimize the production cost while being able to
fulfill the orders.
Decision Variables: Let x1 be the no. of hours operating facility 1
x2 be the no. of hours operating facility 2
x3 be the no. of hours operating facility 3
Objective Function: Minimize Cost, C = 190 x1 + 260 x1 + 150 x3
Constraints:
Orders of Powder Mix 295 x1 + 385 x2 + 350 x3 >= 5,500
Orders for Tea concentrate 260 x1 + 375 x2 >= 4,000
Old facility Operating Hours x1 <=8
Non-negativity constraint: x1 , x2 , x3 >= 0
Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Solve the LP model in Example 1 using Excel Solver

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Solve the LP model in Example 1 using Excel Solver

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Constraints in a Linear Programming Problem

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 2: Portfolio Selection
The International City Rust (ICT) invests in short-term trade credits,
corporate bonds, gold stocks, and construction loans. To encourage a
diversified portfolio, the board of directors has placed limits on the
amount that can be committed to any one type of investment. ICT has
$5 million available for immediate investment and wishes to maximize
the return on the investments made over the next months while
satisfying the diversification requirements as set by the board of
directors.
The board specifies that at least 55% of the funds invested must be
in gold stocks and construction loans and that no less than 15% must
be invested in trade credits.

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 2: Portfolio Selection
The specifics of the investment possibilities are as follows:

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Objective: To maximize the return on investment.
Decision Variables: Let x1 be the amt invested in trade credits
x2 be the amt invested in corporate bonds
x3 be the amt invested in gold stocks
x4 be the amt invested in construction loans
Objective Function: Minimize Returns, Z = 0.07x1+0.11x2+0.19x3+0.15x4
Constraints:
Trade credits investment x1 < = 1,000,000
Corporate bonds investment x2 < = 2,500,000
Gold stocks investment x3 < = 1,500,000
Construction loans investment x4 < = 1,800,000
Available amount x 1 + x2 + x3 + x4 < = 5,000,000
Board specs 1 x3 + x 4 > = 0.55(x1 + x2 + x3 + x4)
Board specs 2 x1 > = 0.15(x1 + x2 + x3 + x4)
Non-negativity constraint: x1 , x2 , x3,, x4 >= 0
Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Solve the LP model in Example 2 using Excel Solver

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 3: Product Mix Problem

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 3: (continued)
Objective: To find the best mix of computer models to attain
maximum profit.
Decision Variables: Let x1 be the no. of units of Basic
x2 be the no. of units of XP
Objective Function: Maximize Profit, Z = 80 x1 + 129 x2
Constraints:
Assembly hours 5 x1 + 6 x2 < = 5,500
Testing hours x1 + 2 x2 < = 3,000
Demand for Basic x1 < = 600
Demand for XP x2 < = 1,200
Non-negativity constraint: x1 , x2 >= 0

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Solve the problem using Excel Solver
Labor cost Labor cost
SP Cost (assembly) (testing) NP
Net profit Basic 300 150 55 15 80
XP 450 225 66 30 129

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
Exercise 1:
 Optimization with Integer Linear Programming
In some optimization problems, it is critical that decision
variables are integers (or whole numbers). In general,
rounding optimal values to the nearest integer leads to
suboptimal solutions. Hence, we now employ integer linear
programming.
Two common types of IP problems are capital budgeting
problems for the selection of investment projects and
transportation problems for the selection of supply locations
for product delivery.

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Types of Integer Linear Programming Models
A. All-Integer Linear Program – an LP problem where all
variables are required to be integers.
Example:

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Types of Integer Linear Programming Models
A. All-Integer Linear Program

If the phrase “integer” is dropped, we have the familiar LP

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Types of Integer Linear Programming Models
B. Mixed-Integer Linear Program – an LP problem where
some, but not necessarily all, are required to be integers.

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Types of Integer Linear Programming Models
C. 0 -1 Integer Linear
Program – an LP
problem where the
integer variables may
only take on the
values 0 or 1.

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 4

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 4: (continued)
Objective: To find the best mix of investment alternatives to
maximize annual cash flow.
Decision Variables: Let x1 be the no. of townhouses to purchase
x2 be the no. of apartment buildings to purchase
Objective Function: Maximize Cash flow, Z = 10 x1 + 15 x2 (in thousands)
Constraints:
Funds Availability ($1000s) 282 x1 + 400 x2 < = 2,000
Manager’s time, hours 4 x1 + 40 x2 < = 140
Townhouses available x1 <=5
Non-negativity constraint: x1 , x2 >= 0 and integer

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 4: (continued)
Solution of problem using Excel Solver.

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 5: (0-1 Integer Linear Programming)
The Ice-Cold Refrigerator Co. is considering investing in
several projects that have varying capital requirements over
the next four years. Faced with limited capital each year,
management would like to select the most profitable projects.
The estimated net present value for each project, the capital
requirements, and the available capital over the four-year
period are shown in the table on the next slide.

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 5: (0-1 Integer Linear Programming)

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 5: (0-1 Integer Linear Programming)

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 4: (0-1 Integer Linear Programming)

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Example 4: (continued)
Solution of problem using Excel Solver.
Objective Function

Maximize NPV of the capital budgeting projects z 140000 The company should venture into the following projects:
Plant expansion, warehouse expansion and new machinery project
Decision Variables to realize a maximum NPV of $140,000.
Plant expansion project 90000 P 1
Warehouse expansion project 40000 W 1
New Machinery project 10000 M 1
New product research project 37000 R 0

Constraints
Year 1 capital available 35 <= 40
Year 2 capital available 35 <= 50
Year 3 capital available 40 <= 40
Year 4 capital available 24 <= 35

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Practice Problems
(Formulate the LP model and solve using Excel Solver add-in. )

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Practice Problems
(Formulate the LP model and solve using Excel Solver add-in. )

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Practice Problems
Solve using Excel Solver add-in.

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
 Practice Problems
(Formulate the LP model and solve using Excel Solver add-in. )

Business Analytics: Data analysis and decision-making by Aljbright and Winston, 2020
References
• Business Analytics by Jaggia et al (2020)
• Business Analytics: Data analysis and decision-making by Albright and
Winston
• Introduction to Management Science by Taylor (2012)
• Quantitative Approaches to Management by Render et al (2018)