NAME RUTVIK ASHOKKUMAR WADKAR

ROLL NUMBER 2314518864

SESSION FEBRUARY - MARCH 2024
PROGRAM MASTER OF BUSINESS ADMINISTRATION (MBA)
SEMESTER II
COURSE CODE & DMBA205- OPERATIONS RESEARCH
NAME

ANSWERS SET - 1

Q1. What is Operations Research? Explain the Methodology used to solve Operations
Research Problems in brief.
A1.
Definition: Meaning of Operations Research
Churchman, Ackoff, and Aruoff described operations research as "the application of scientific
methods, techniques, and tools to the operation of a system to find the optimal solutions to
problems," with 'optimal' meaning the best possible alternative.
The goal of operations research (OR) is to offer a scientific foundation for decision-makers to
solve issues involving various organizational components. This is accomplished by a
multidisciplinary team of scientists collaborating to identify the best solution for the
organization as a whole. This solution is known as an optimal decision.
Operations research can also be defined as "the use of scientific methods to establish criteria
for decisions involving people, machines, and systems with repetitive operations."
The Methodology used to solve Operations Research Problems
The fundamental characteristic of operations research (OR) is its use of mathematical models to
analyze problems. This unique approach adapts the scientific methodology from the physical
sciences. The scientific method involves translating a given problem into a mathematical
representation, solving it, and then interpreting the solution in the original context. Figure 1.3
illustrates the OR approach to problem-solving, which involves the following steps: defining the
problem, constructing the model, solving the model, validating the model, and implementing the
results.
1. Problem Definition
The first and most crucial step in the OR approach to problem-solving is defining the problem.
Proper identification of the problem is essential because the problem statement will highlight the
following three major aspects:
 Description of the goal or objective of the study
  Identification of the decision alternatives available to the system
 Recognition of the limitations, restrictions, and requirements of the system

2. Model Construction
Based on the problem definition, the next step is to identify and select the most suitable model to
represent the system. It is crucial to ensure that the chosen model specifies quantitative expressions
for the objective and constraints of the problem in terms of its decision variables. A model provides
a comprehensive view of the entire problem, facilitating a structured approach to addressing it. If
the resulting model aligns with common mathematical models, mathematical techniques can be
used to find a convenient solution. However, if the mathematical relationships within the model
are too complex for analytical solutions, a simulation model might be more appropriate. Therefore,
constructing the appropriate model is essential.

3. Model Solution
Once an appropriate model is selected, the next step is to develop a solution and interpret it within
the context of the given problem. Solving the model involves determining a specific set of decision
variables that yield the optimal solution. An optimal solution is one that either maximizes or
minimizes the performance of a particular measure within the model, subject to the conditions and
constraints imposed.
4. Model Validation
A model serves as a good representation of a system, but its optimal solution must enhance the
system's performance. To test a model's validity, compare its performance against historical data
from the actual system. If the model can replicate past performance under similar input conditions,
it is likely valid. However, this does not guarantee that future performance will match past
behavior. Additionally, since the model relies on thorough analysis of historical data, comparisons
should generally yield positive results. In some cases, this issue can be addressed by using data
from trial runs of the system. It's important to note that such validation methods are unsuitable for
non-existent systems, as there is no data available for comparison.
5. Result Implementation
The optimal solution derived from the model must be implemented within the system to observe
any performance improvements. This performance evaluation should be validated under varying
conditions. To achieve this, the results must be translated into clear and detailed operating
instructions for those who will manage and operate the recommended system. During this phase,
the interaction between the operations research team and the operating personnel is at its most
intensive.
Q2. Solve the following linear programming problem using its Dual form:
Minimize Z = 3x1 + 4x2
Subject to: 4x1 + x2 ≥ 30
-x1 - x2 ≤ -18
x1 +3x2 ≥ 28
where x1, x2 ≥ 0
A.2.
Primal Problem
Minimize Z=3x1+4x2
Subject to:

Dual Problem

Maximize W = 30y1+18y2+28y3

Subject to:

4y1−y2+y3≤3
y1−y2+3y3 ≤4
y1,y2,y3≥0

We will use the Simplex method to solve the dual problem. First, we write down the initial simplex
tableau for the dual problem.

Initial Simplex Tableau

Basis B Y1 Y2 Y3 Solution
Y4 0 4 -1 1 3
 Y5 0 1 -1 3 4
z-w -30 -18 -28

Here, y4 and y5 are the slack variables introduced to convert inequalities into equalities.
Now, we can solve these equations to find the values of 𝑦1, 𝑦2, and 𝑦3.
To solve the dual problem, we first convert the inequalities to equations:
4𝑦1 − 𝑦2 + 𝑦3 = 3
𝑦1 + 3𝑦2 = 4
Now, we can solve these equations to find the values of 𝑦 , 𝑦 , and 𝑦 .
From the first equation, we have:
4𝑦1 − 𝑦2 + 𝑦3 = 3 ⟹ 4𝑦1 = 𝑦2 − 𝑦3 + 3

Substituting 𝑦1 = into the second equation, we get:

𝑦2 − 𝑦3 + 3
+ 3𝑦2 = 4 ⟹ 𝑦2 − 𝑦3 + 3 + 12𝑦2 = 16 ⟹ 13𝑦2 − 𝑦3 = 13
4
Solving these two equations simultaneously, we get:
13𝑦2 − 𝑦3 = 13
4𝑦1 = 𝑦2 − 𝑦3 + 3

Now, we substitute 𝑦2 = into the primal constraint 𝑦1 + 3𝑦2 ≥ 4 to find the value of 𝑦1 :

𝑦3 + 13 52 − 3𝑦3 − 39 13 − 3𝑦3
𝑦1 + 3 ≥ 4 ⟹ 𝑦1 ≥ ⟹ 𝑦1 ≥
13 13 13
The dual problem is maximized when 𝑦1 takes the minimum value that satisfies the constraint.
Since 𝑦1 ≥ , the minimum value of 𝑦1 occurs when 𝑦3 = 0.

Substituting 𝑦3 = 0 into the equations, we get 𝑦2 = 1 and 𝑦1 = = 1.

Optimal Solution for Dual Problem

 y1=1.36
 y2=0
 y3=1.18
Optimal Value : W=30(1.36)+28(1.18)=40.8+33.04=73.84
The optimal value of the dual problem is 73.84. The corresponding primal solution (shadow prices)
can be found by solving the primal problem with these dual values. However, the optimal values
of the primal variables x1 and x2 need further verification by re-solving using simplex or another
method for confirmation.
Q3. A firm marketing a product has four salesman S1, S2, S3 and S4. There are three
customers to whom a sale of each unit to be made. The chance of making a sale to a
customer depend on the salesman customer support. The data depicts the probability with
which each of the salesman can sell to each of the customers.

Salesman
Customer S1 S2 S3 S4
C1 0.7 0.4 0.5 0.8
C2 0.5 0.8 0.6 0.7
C3 0.3 0.9 0.6 0.2

If only one salesman is to be assigned to each of the customers, what combination of

salesman and customers shall be optimal. Give further that the profit obtained by selling
one unit of C1 is Rs. 500, whereas it is respectively Rs 450 and Rs. 540 for sale to C2 and
C3. What is the expected profit?
A.3.
Introducing a dummy row and applying Hungarian method you have:
Salesman Profit
Customer S1 S2 S3 S4
C1 0.7 0.4 0.5 0.8 500
C2 0.5 0.8 0.6 0.7 450
C3 0.3 0.9 0.6 0.2 540
C4 00 00 00 00 000
Total
Profit

Multiply each customers profit value with the probability of each salesman and customer
Customer Salesmen
S1 S2 S3 S4
C1 0.7*500 0.4*500 0.5*500 0.8*500
C2 0.5*450 0.8*450 0.6*450 0.7*450
C3 0.3*540 0.9*540 0.6*540 0.2*540
We have to assign on salesmen to one customer
Since there are four salesman and three customers we have to add one customer
(dummy cell) So the matrix (table) will be as follow:

Customer Salesmen
S1 S2 S3 S4
C1 350 200 250 400
C2 225 360 270 315
C3 162 486 324 108
C4 0 0 0 0

Row Deduction
Customer Salesmen
S1 S2 S3 S4
C1 0
150 50 200
C2
0 135 45 90
C3
54 378 216 0
C4
0 0 0 0

Optimal Assignment
C1 to S2 =200
C2 to S1 =225
C3 to S4 =108
C4 to S3 =000
Total Expected Profit is Rs. 533
 Assignment Set - 2

Q4. What is Monte Carlo simulation? Explain Monte Carlo Simulation Procedure in
brief.

A.4 Monte Carlo Simulation: A Powerful Tool for Decision Making

Monte Carlo simulation is a computational method used to evaluate the effects of risk and
uncertainty in decision-making processes. Named after the renowned Monte Carlo Casino in
Monaco, famous for its games of chance, this technique uses random sampling and probability
distributions to simulate various potential outcomes in a given problem. It offers a spectrum of
possible results and the likelihood of each occurring, aiding decision-makers in quantifying
risk and making more informed choices.
The Monte Carlo simulation procedure involves several key steps:

1. Problem Formulation: The first step is to clearly define the problem and the variables
involved. This includes identifying the decision to be made, the uncertain factors affecting the
decision, and the range of possible outcomes.

2. Model Development: Next, a mathematical model is developed to represent the problem.

This model includes the relationships between the variables and the probability distributions
that describe the uncertainty in the system.

3. Random Sampling: Monte Carlo simulation relies on random sampling to generate possible
outcomes. Random numbers are generated based on the specified probability distributions for
each variable.

4. Simulation Runs: The simulation is run multiple times, with each run representing a different
possible outcome. The number of runs is typically large enough to ensure a reliable estimate
of the outcomes.

5. Analysis of Results: Once the simulation runs are completed, the results are analyzed to
determine the range of possible outcomes and their probabilities. This analysis provides
insights into the risk and uncertainty associated with the decision.

6. Decision Making: Finally, based on the results of the simulation, a decision is made. The
decision-maker can use the information provided by the simulation to assess the risks and
benefits of different options and make an informed choice.
Example of Monte Carlo Simulation:

Consider a project manager who needs to estimate the completion time for a construction
project. Several uncertain factors could affect the project's timeline, such as weather conditions,
availability of resources, and unforeseen delays. By using Monte Carlo simulation, the project
manager can model these factors using probability distributions and generate multiple possible
scenarios for the project's completion time.
Based on the simulation results, the project manager can assess the likelihood of completing
the project on time and within budget, and identify potential risks. This information can help
in making decisions such as resource allocation, scheduling, and risk management strategies.
Monte Carlo simulation is a powerful tool for decision-making in complex and uncertain
situations. It provides a systematic way to assess risks and uncertainties, helping decision-
makers make more informed choices. By simulating different scenarios and analyzing the
results, Monte Carlo simulation enables organizations to optimize their decision-making
processes and improve their outcomes.
Answer 5:
Making use of time estimates t , t and t , the calculations for expected time t and variance
σ for activities are shown in table below:

Activity Time Duration (Weeks)

𝟐
𝐭𝐩 − 𝐭𝐨
𝐭 𝐨 + 𝟒𝐭 𝐦 + 𝐭 𝐩 𝛔 =𝟐
Sequence 𝟔
𝐭𝐨 𝐭𝐦

1−2 1 1 7 2 1

1−3 1 4 7 4 1

1−4 2 2 8 3 1

2−5 1 1 1 1 0

3−5 2 5 14 6 4

4−6 2 5 8 5 1

5−6 3 6 15 7 4
 1−4 2 2 8 3 0

2−5 1 1 1 1 4

3−5 2 5 14 6 1

4−6 2 5 8 5 4

5−6 3 6 15 7

1. The network diagram can be drawn as follows:

Network Diagram

The evpecned duration and variance of each activity is given in table above. We now calculate
the expected
 Variwes Paths Length of the Path

1−2−5−6 1−2−5−6=2+1+7 = 10 weeks

1−3−5−6 1−3−5−6=4+6+7 = 17 weeks

1−4−6 1−4−6= 3+5 = 8 weeks

Among these putis, the critical path is 1 − 3 − 5 − 6 = 17 weeks

2. Wuriance of the project length is the sum of the variances of the activities on the critical
path. Hence, Variance,

σ = 1 + 4 + 4 = 9 deys (trom table above). Therefore, Standard deviation of the project σ =

√9 = 3 weeks

Phobubiliry that the project will be completed atleast 4 weeks earlier than expected time is
given by;

T −T 13 − 17
ℙ(x < 13) = ℙ ℤ < =ℙ Z< = ℙ[Z < −1.33] = 1 − 0.9082
σ 3
= 0.0918

Answer – 6
To solve this, let's first construct the payoff matrix:

Let's represent the coins as 1, 2, 5, 10, and 50 rupees, respectively. Player A is the maximizing
player and Player B is the minimizing player. The possible totals can be odd (win for A) or
even (loss for A). The matrix will have rows for Player A's choices and columns for Player B's
choices, and the entries will represent the payoff to A.

Here's the payoff matrix:

A\B 1 2 5 10 50
---- --- --- --- ---- ----
1 -1 1 -1 1 -1
2 1 -1 1 -1 1
5 -1 1 -1 1 -1
10 1 -1 1 -1 1
 50 -1 1 -1 1 -1

In this matrix:

 A positive number represents a win for A (A gains B's coin).

 A negative number represents a loss for A (A loses their coin to B).

 Zero represents no change in coins.

To find the optimal strategies for each player, we can use the concept of mixed strategies. Let
𝑝p be the probability that A chooses 1 rupee, 𝑞q be the probability that A chooses 2 rupees, 𝑟r
be the probability that A chooses 5 rupees, 𝑠s be the probability that A chooses 10 rupees, and
𝑡t be the probability that A chooses 50 rupees.

Let's assume B's optimal strategy is to randomize their choices such that all coins are equally
likely to be chosen. This assumption simplifies the calculation, and it's often the case in such
games where one player is maximizing and the other is minimizing.

The expected payoff for A when choosing 1 rupee is:

𝐸1=𝑝(−1)+𝑞(1)+𝑟(−1)+𝑠(1)+𝑡(1)=𝑞+𝑠+𝑡−𝑝−𝑟

Similarly, the expected payoffs for A when choosing 2, 5, 10, and 50 rupees are:

𝐸2=𝑝(1)+𝑞(−1)+𝑟(1)+𝑠(−1)+𝑡(−1)=𝑝+𝑟−𝑞−𝑠−𝑡

𝐸5=𝑝(−1)+𝑞(1)+𝑟(−1)+𝑠(1)+𝑡(−1)=𝑞+𝑠−𝑝+𝑟−𝑡

𝐸10=𝑝(1)+𝑞(−1)+𝑟(1)+𝑠(−1)+𝑡(1)=𝑝+𝑟−𝑞−𝑠+𝑡

𝐸50=𝑝(−1)+𝑞(1)+𝑟(−1)+𝑠(1)+𝑡(0)=𝑞+𝑠−𝑝+𝑟

To find the optimal strategy for A, we need to find the values of p, q, r, s, and t that maximize
A's expected payoff while considering the constraints p+q+r+s+t=1 and 𝑝,𝑞,𝑟,𝑠,𝑡≥0

The optimal strategy for B is to choose all coins with equal probability, i.e., 𝑞=𝑟=𝑠=𝑡=0.2

By solving the above system of equations, we can find the optimal strategies for A and the
value of the game to A.

Let's solve for the optimal strategies for A and the value of the game to A.
1. Optimal Strategy for A:

We want to maximize E1, E2, E5, E10, and E50 subject to the constraint 𝑝+𝑞+𝑟+𝑠+𝑡=1 and
𝑝,𝑞,𝑟,𝑠,𝑡≥0

Let's set up the equations:

 𝐸1=𝑞+𝑠+𝑡−𝑝−𝑟

 𝐸2=𝑝+𝑟−𝑞−𝑠−𝑡

 𝐸5=𝑞+𝑠−𝑝+𝑟−𝑡

 𝐸10=𝑝+𝑟−𝑞−𝑠+𝑡

 𝐸50=𝑞+𝑠−𝑝+𝑟

 𝑝+𝑞+𝑟+𝑠+𝑡=1

2. Solving the System of Equations:

 Adding all equations, we get: 2𝑝+2𝑞+2𝑟+2𝑠+2𝑡=2

 Dividing by 2, we get: 𝑝+𝑞+𝑟+𝑠+𝑡=1 which is our constraint.

 Therefore, the equations are consistent with the constraint.

3. Optimal Strategies:

 The equations are symmetrical, so any permutation of the solutions is also valid.

 One possible optimal strategy for A is p=0.2, q=0.2, r=0.2, s=0.2, 𝑡=0.2t=0.2.

4. Value of the Game to A:

Using the optimal strategy, the expected payoff to A is

𝐸total=0.2×𝐸1+0.2×𝐸2+0.2×𝐸5+0.2×𝐸10+0.2×𝐸50

Calculating Etotal, we get the value of the game to A.

Let's calculate 𝐸totalEtotal to find the value of the game to A.

Substituting the optimal strategy p=0.2, q=0.2, r=0.2, s=0.2, t=0.2 into the expressions for E1,
E2, E5, E10, and E50:

𝐸1=0.2(−1)+0.2(1)+0.2(−1)+0.2(1)+0.2(1)=0
𝐸2=0.2(1)+0.2(−1)+0.2(1)+0.2(−1)+0.2(−1)=−0.2

𝐸5=0.2(−1)+0.2(1)+0.2(−1)+0.2(1)+0.2(−1)=−0.2

𝐸10=0.2(1)+0.2(−1)+0.2(1)+0.2(−1)+0.2(1)=0.2

𝐸50=0.2(−1)+0.2(1)+0.2(−1)+0.2(1)+0.2(0)=0

Now, calculating the total expected payoff to A:

𝐸total=0.2×𝐸1+0.2×𝐸2+0.2×𝐸5+0.2×𝐸10+0.2×𝐸50

𝐸total=0×0.2+(−0.2)×0.2+(−0.2)×0.2+0.2×0.2+0×0.2

𝐸total=0−0.04−0.04+0.04+0

𝐸total=−0.08

So, the value of the game to Player A is -0.08, which means, on average, Player A loses 0.08
rupees per game against an optimal strategy from Player B.