Business Analytics Lecture 1

Lecture 6

 Framework for Business Analytics

 Bland Brewery Linear Programming Model

 Graphical representation of linear programs

 Integer Programming
Business Analytics Lecture 1 2

Business Intelligence and Analytics

Prescriptive How do you make it happen?

Optimization, Planning

Predictive What will happen?

Forecasts, Machine learning, Simulation
Value

Diagnostic
Why did it happen?
Queries, Data mining, Statistical Analysis

Descriptive
What Happened?
Reports, Alerts, Mapping

Difficulty
Business Analytics Lecture 1 3

Business Intelligence and Analytics

Prescriptive How do you make it happen?

Optimization, Planning

Predictive What will happen?

Forecasts, Machine learning, Simulation
Value

Diagnostic
Why did it happen?
Queries, Data mining, Statistical Analysis

Descriptive
What Happened?
Reports, Alerts, Mapping

Difficulty
Business Analytics Lecture 1 4

Business Intelligence and Analytics

Prescriptive How do you make it happen?

Optimization, Planning

Predictive What will happen?

Forecasts, Machine learning, Simulation
Value

Diagnostic
Why did it happen?
Queries, Data mining, Statistical Analysis

Descriptive
What Happened?
Reports, Alerts, Mapping

Difficulty
Business Analytics Lecture 5

Bland Brewery Problem

“Raw” Components Products

Corn Beer
Hops
Malt
Ale
Business Analytics Lecture 6

Bland Brewery Problem (Continued)

• Profitability

1 Barrel of Beer 1 Barrel of Ale

$23 $13

• Mixing Quantities • Availability

1 Barrel of Beer 1 Barrel of Ale Corn 480 lbs

Corn 15 lbs 5 lbs Hops 160 ozs

Hops 4 ozs 4 ozs Malt 1,190 lbs

Malt 20 lbs 35 lbs

How much of beer and ale to produce, if we want to

maximize profits?
Business Analytics Lecture 7

Plan I: Beer-only Production

• Profitability • Mixing Quantities • Availability

1 Barrel of Beer 1 Barrel of Corn 480 lbs

Beer
$23 Corn 15 lbs Hops 160 ozs

Hops 4 ozs Malt 1,190 lbs

Malt 20 lbs

• We run out of corn after producing 480/15=32 barrels of beer

• We run out of hops after producing 160/4=40 barrels of beer

• We run out of malt after producing 1190/20=59.5 barrels of beer

• We can produce at most 32 barrels of beer, making total profit of

$23×32=$736
Business Analytics Lecture 8

Plan II: Ale-only Production

• Profitability • Mixing Quantities • Availability

1 Barrel of Ale 1 Barrel of Ale Corn 480 lbs

$13 Corn 5 lbs Hops 160 ozs

Hops 4 ozs Malt 1,190 lbs

Malt 35 lbs

• We run out of corn after producing 480/5=96 barrels of ale

• We run out of hops after producing 160/4=40 barrels of ale

• We run out of malt after producing 1190/35=34 barrels of ale

• We can produce at most 34 barrels of ale, making total profit of

$13×34=$442
Business Analytics Lecture 9

Can we do better by producing both ale and beer?

• Mixing Quantities

1 Barrel of Beer 1 Barrel of Ale

Corn 15 lbs 5 lbs

Hops 4 ozs 4 ozs

Malt 20 lbs 35 lbs

• Ale and beer are “complements”, we have a chance of

extracting more revenue if we “mix” them in the production
plan
Business Analytics Lecture 10

Bland Brewery Data

• Profitability
1 Barrel of Beer 1 Barrel of Ale

$23 $13

• Mixing Quantities • Availability

1 Barrel of Beer 1 Barrel of Ale Corn 480 lbs

Corn 15 lbs 5 lbs Hops 160 ozs

Hops 4 ozs 4 ozs Malt 1,190 lbs

Malt 20 lbs 35 lbs

Business Analytics Lecture 11

Bland Brewery Model: Standard Notation

 Decision Variables
Let A = # of barrels of ale to produce, and
B = # of barrels of beer to produce.
Note: Use suggestive (mnemonic) variable names for
readability

 Objective Function
Profit in $ = 13A + 23B

 Constraints
Corn Availability: 5A + 15B ≤ 480
Hops Availability: 4A + 4B ≤ 160
Malt Availability: 35A + 20B ≤ 1190
Non-negativity: A, B ≥ 0
Business Analytics Lecture 12

Bland Brewery Linear Program

max 13 A + 23 B (Profit) Objective Function

subject to Coefficients
(corn) 5A + 15B ≤ 480
(hops) 4A + 4B ≤ 160
Right hand sides
(malt) 35A + 20B ≤ 1190
(nonnegativity) A, B ≥ 0

Variables
Business Analytics Lecture 13

Figure 1. The preliminary spreadsheet BLAND.XLS

Business Analytics Lecture 14
Cell F5
=SUMPRODUCT(C4:D4, C5:D5)

Cell E10
=SUMPRODUCT($C$4:$D$4, C10:D10)

Figure 2. The spreadsheet BLAND.XLS with formulas

Business Analytics Lecture 15

Figure 3. The Solver Parameters dialog box

A description of the Excel spreadsheet optimizer is given in the reading

“An introduction to Spreadsheet Optimization using Excel”.
Business Analytics Lecture 16

Figure 4. The Solver Parameters dialog box with constraints

added
Business Analytics Lecture 17

Figure 5. The Solver Options dialog box

Business Analytics Lecture 18

Figure 6. The spreadsheet after optimizing

Business Analytics Lecture 19

Cell F10
=IF(E10<=G10+0.00001, “<=“, “NOT <=“)

Figure 7. The spreadsheet with constraints indicated

Business Analytics Lecture 20

Terminology

 Feasible and Infeasible Solutions

 A production plan (A,B) that satisfies all of the constraints is

called a feasible solution

 For example, in the Bland Brewery LP, the solution (A=10,

B=10) is feasible.

 Constraints
Corn Availability: 5 x 10 + 15 x 10 = 200 ≤ 480
Hops Availability: 4 x 10 + 4 x 10 = 80 ≤ 160
Malt Availability: 35 x 10 + 20 x 10 = 550 ≤ 1190
Non-negativity: 10, 10 ≥ 0
Business Analytics Lecture 21

Terminology (continued)

 Feasible and Infeasible Solutions

 The production plan (A=40, B=10) is not feasible, i.e. it is

infeasible because the hops and malt constraints are violated

 Constraints
Corn Availability: 5 x 40 + 15 x 10 = 350 ≤ 480
Hops Availability: 4 x 40 + 4 x 10 = 200 > 160
Malt Availability: 35 x 40 + 20 x 10 = 1600 > 1190
Non-negativity: 40, 10 ≥ 0
Business Analytics Lecture 22

 Optimal Solution
 For a maximization (respectively, minimization) problem, an
optimal solution is a feasible solution that has the largest
(respectively, smallest) objective function value among all
feasible solutions

 The optimal solution for the Bland Brewery production model

is (A=12, B=28). The optimal objective function value is $800.
Business Analytics Lecture 23

Assumptions in a Linear Program

 Continuity: the decision variables are continuous, i.e., fractional

values are allowed

 Proportionality: for example, it takes twice as much hops to

make twice as much beer or ale; there are no economies of
scale

 Additivity: profit is the sum of the profit contributions from ale

and beer
Business Analytics Lecture 24

Assumptions in a Linear Program

 In short, the objective function and constraints must be linear

with respect to decision variables

 Linear functions of A and B:

 13A + 23B
 0.5A + (2/3)B

 Non-Linear functions of A and B:

 13A2 + 23AB
 log(A) + cos(B)
 max(A,0)
 IF(A< 5,0,10)
Business Analytics Lecture 25

Assumptions in a Linear Program

.
Allowable variations:

 Objective function can be maximized or minimized

 Constraints can be ≥, ≤, or =

 Noninteger or integer coefficients and right-hand sides are

allowed

 Negative or positive coefficients and right-hand sides are allowed