The Nature of Management Science

Management Science
- Discipline that attempts to aid managerial decision making by applying a scientific
approach to managerial problems that involve quantitative factors
- Operations Research: traditional name given to the discipline
- Business analysts employing management science don’t make managerial decisions,
managers do.

Decision Making
● Managers’ responsibility:
- To make strategic, tactical, or operational decisions
● Strategic Decisions:
- 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
- Usually the responsibility of mid level 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

Common Approaches to Making Decisions

1. Tradition
2. Intuition
3. Rules of Thumb
4. Using the Relevant Data Available

Steps in Decision Making Process

1. Define the problem and gather data
2. Determine criteria to evaluate alternative solutions
3. Formulate a model to represent the problem
4. Develop a computer based procedure for deriving solutions to the problem
5. Test the model and refine if needed
6. Apply to model to analyze the problem
7. Management chooses an alternative
8. Help implement the team’s recommendations that are adopted by the management

Business Analytics
- Defined as the art of transforming data into insight for making better business decisions
 - Big data: massive amounts of data are now commonly available that help guide
managerial decision making
- Primary focus: how to make the most effective use of all these data

Descriptive Analytics: Data visualization

Predictive analytics: Applying statistical models to predict future events
Prescriptive Analytics: Prescribe what should be done in the future
- Used are powerful optimization models and techniques of management science

The Role of Data Science

- Interdisciplinary field that uses scientific methods, processes, algorithms, and systems to
extract knowledge or insights from even massive amounts of data
- More on scientific methods and more applicable to various areas in beyond business

The Role of Machine Learning

- Allow computers to learn automatically from historical relationships and trends in the data
in order to do such things as making data-driven predictions

The Role of Artificial Intelligence

- AI is to build intelligent computer programs and machines that can stimulate human
thinking capability and behavior
- Machine Learning (ML) and AI are closely related technologies
- ML is part of AI

The Relationship between Management Science and Business Analytics

- They complement each other extremely well
- Someone who once was a management science specialist now needs to be well trained
in business analytics, vice versa
- Business schools now need to train future business analysts who have a solid foundation
in the traditional disciplines of both management and business analytics

Performing Descriptive Analytics to Explore the Data

- Descriptive analytics needs to be performed to explore the data to better understand the
relationship between advertising and sales
- One main technique is called data visualization: scatter plot

Linear Regression
- Involved approximating the relationship between the dependent variable and the
independent variable by a straight trendline
- Formula: y=mx+b
 Y = estimated value of the dependent variable
A = slope of the linear regression line
X = value of the independent variable
B = intercept of the linear regression trendline with the y-axis

Decision Analysis

Managers often must make decisions in environments that are fraught with uncertainty
Decision Analysis
- Can be used to develop an optimal strategy
- Also known as decision theory: general approach that helps decision makers make
intelligent choices
- A good decision analysis includes careful consideration of risk
- Considers problems that involve reasonably few decision alternatives and reasonably few
possible future events
Risk Analysis
- Helps to provide the probability information about the favorable as well as the
unfavorable outcomes that may occur

Terminologies
Term Meaning

Decision Maker Individual or group responsible for making the decision

Decision Alternatives Options for the decision to be made

Chance Events Random factors outside the control of the decision maker.
These determine the situation that will be found when the
decision is executed

State of Nature Each of the possible situations (only one state of nature can
occur)

Prior Probabilities The decision maker generally will have some information
about the relative likelihood of the possible states of nature

Payoff Quantitative measure of the value to the decision maker of the

outcome (monetary value).

Payoff Table Table showing payoffs for all combinations of decision

alternatives and states of nature
Decision Criterion
1. Certainty: we know for certain which state of nature will occur
2. Uncertainty: we have no information on the likelihoods of various states of nature
3. Risk: the likelihood (probability) of each state of nature can be estimated
>>> It is unlikely that we know for certain which state of nature will actually occur, we simply
choose the alternative that gives the best payoff for that state of nature.

Decision Analysis Under Uncertainty

● Best-case and worst-case analysis is sufficient
● Where the decision maker has little confidence in their ability to assess the probabilities
1. Maximax: optimistic approach
- Focuses only on the best that can happen
- Steps: (1) identify the maximum payoff from any state of nature, (2) find the
maximum payoffs between the chosen ones in the first step
2. Maximin: conservative approach
- Focuses only on the worst that can happen
- Steps: (1) identify the minimum payoff from any state of nature, (2) find the
maximum payoffs between the chosen ones in the first step
3. Minimax Regret
- Opportunity loss: regret
- One would choose the decision alternative that minimizes the maximum
state of regret that could occur over all possible states of nature
- Steps: (1) compute the regret (subtract each entry in a column from the
largest entry in the column) (2) determine the maximum regret for each
alternative (3) find the minimum of these maximum regrets and choose this
alternative

Linear Programming

Optimization Problems
- Can be used to support and improve managerial decision making
- Maximize or minimize some function (objective function) and have a set of restrictions
(constraints)
- Can be linear or nonlinear

Term Meaning

Problem Formulation Process of translating the verbal statement of a problem into a

mathematical statement

Objective Function Mathematical expression that gives the measure of

 performance for the problem in terms of decision variables

Decision Variables/ Quantity whose value is assigned by the decision maker.

Controllable Values

Parameters Constants in the mathematical model

- Uncontrollable inputs

Constraint Inequality or equation that expression that expresses some

restrictions on the values that can be assigned to the decision
variables

Linear Functions
- Linear optimization models are also known as linear programs
- Planning of activities represented by a linear mathematical model
- Linear: feature of the objective function and the constraints
- Linear Function/Model: mathematical function in which each variable appears in a
separate term and is raised to the first power
Properties of Linear Function
1. Additivity: contribution from one decision gets added to contributions of other decisions
2. Proportionality: contribution from a decision grows in proportion to the value of the
decision variable
3. Divisibility: can have fractional decision values
Feasible solution
- a setting of a decision variables that satisfies all of the constraints of the problem
- Feasible region: a set of all possible solutions

Linear Programming: Formulation and Applications

Resource-Allocation Linear programming problems involving the allocation of resources

to activities
- Identifying feature of any such problem is that each
functional constraint in the LP model is a resource constraint
𝑎𝑚𝑜𝑢𝑛𝑡 𝑜𝑓 𝑟𝑒𝑠𝑜𝑢𝑟𝑐𝑒 𝑢𝑠𝑒𝑑 ≤ 𝑎𝑚𝑜𝑢𝑛𝑡 𝑜𝑓 𝑟𝑒𝑠𝑜𝑢𝑟𝑐𝑒 𝑎𝑣𝑎𝑖𝑙𝑎𝑏𝑙𝑒

Cost Benefit Trade-Off Linear programming problems where the mix of levels of various
activities is chosen to achieve minimum acceptable levels of various
benefits at a minimum cost
𝑙𝑒𝑣𝑒𝑙 𝑎𝑐ℎ𝑖𝑒𝑣𝑒𝑑 ≥ 𝑚𝑖𝑛𝑖𝑚𝑢𝑚 𝑎𝑐𝑐𝑒𝑝𝑡𝑎𝑏𝑙𝑒 𝑙𝑒𝑣𝑒𝑙

Fixed-Requirement Linear programming problems where activities are provided to meet

requirements (Ex. transportation and assignment problems)
𝑎𝑚𝑜𝑢𝑛𝑡 𝑝𝑟𝑜𝑣𝑖𝑑𝑒𝑑 = 𝑎𝑚𝑜𝑢𝑛𝑡 𝑟𝑒𝑞𝑢𝑖𝑟𝑒𝑑
 Transportation One of the most common application of linear programming involves
optimizing a shipping plan for transporting goods
- In a typical application, a company has several plants
(sources) producing a certain product that needs to be
shipped to the company’s customers (destinations)

Assignment Special type of linear programming problem

- Involves making assignments, such as assignment of people
to jobs, assigning machines, vehicles or plants to tasks
- Involves binary variables

Mixed Problems Problems that contain more than one type of constraint (mixed
constraints)

Blending Model: Integer Linear Optimization Model

- A variation to the linear programming problem that has the additional restriction that
some or all of the decision variables must have integer values
- Blending problems arise when a decision must be made regarding the blending of two or
more resources to produce one or more products
1. All-integer linear program: all variables are required to be integer
2. Mixed-integer linear program: if some, but not necessarily all variables are required to be
integer
3. Binary integer linear program: the integer variables may take on only the values of 0 and
1

>> Divisibility Assumption of Linear Programming: decision variables in a linear programming

model are allowed to have any values, including fractional values, that satisfy the functional and
nonnegativity constraints. Thus, these variables are not restricted to just integer values

>> The linear program that results from dropping the integer requirements is called the linear
programming relaxation (LP relaxation) of the integer linear program (feasible region)

>> In many cases, a noninteger solution can be rounded to obtain an acceptable integer solution.
However, rounding may not always be a good strategy; when the decision variables take on
small values that have a major impact on the value of the objective function, an optimal integer
solution is needed

BInary Integer Programming

- One of the three key assumptions of linear programming is the continuous-variables
assumption, which requires that the decision variables can take on any nonnegative
values (not just integer values) that fall within its feasible range
 - A problem that fits linear programming except that some or all of its decision variables are
required to have integer values (integer programming problem)
- If only some of the decision variables have this restriction, the problem is referred to as
mixed integer problem

Binary Decision Variables

- Variables that can have only 0,1
- They are widely used in integer programming models to represent yes-or-no decisions,
where a value of 1 corresponds to a yes decision and a value of 0 corresponds to a no
decision (binary integer programming models BIP)

Nonlinear Optimization Problem

- Any optimization problem in which at least one term in the objective function or a
constraint is nonlinear
- Nonlinear programming is used to model nonproportional relationships between activity
levels and the overall measure of performance

Proportionality Assumption of Non-linear Programming

- When any activity has a non proportional relationship where the contribution of the
activity to the measure of performance is not proportional to the level of activity

Local and Global Optima

Local Optima:
- Feasible solution when there are no other feasible solutions with a better objective
function value in the immediate neighborhood
- May be either a local maximum or a local minimum
Local Maximum:
- Feasible solution when there are no other feasible solutions with a larger
objective function value
Local Minimum:
- Feasible solution when there are no other feasible solutions with a
smaller objective function value
Global Optimum:
- Feasible solution when there are no other feasible solutions with a better objective
function value in the entire feasible region
- May either be a global maximum or a global minimum
Global Maximum:
- Feasible solution when there are no other feasible solutions with a larger
objective function value in the entire feasible region
- Also a local maximum
Global Minimum:
 - Feasible solution when there are no other feasible solutions with a
smaller objective function value in the entire feasible region
- Also a local minimum
2 2
Concave Function: function that is bowl-shaped down (− 𝑥 − 𝑦 )
- Maximum value = (0,0)
- Optimal Value = 0
- No larger than 0
- If all squared terms in a quadratic function have a negative coefficient and there are no
cross-product terms
- Only has mountain
2 2
Convex Function: function that is bowl-shaped up (𝑥 + 𝑦 )
- Minimum value = 0
- No less than 0
- Local minimum = global minimum
Queueing

Arrivals
- The time between consecutive arrivals to a queueing system are called interarrival times
- The expected number of arrivals per unit time is referred to as the mean arrival rate
- The symbol used for the mean arrival rate is
λ = 𝑚𝑒𝑎𝑛 𝑎𝑟𝑟𝑖𝑣𝑎𝑙 𝑟𝑎𝑡𝑒 𝑓𝑜𝑟 𝑐𝑢𝑠𝑡𝑜𝑚𝑒𝑟𝑠 𝑐𝑜𝑚𝑖𝑛𝑔 𝑡𝑜 𝑡ℎ𝑒 𝑞𝑢𝑒𝑢𝑒𝑖𝑛𝑔 𝑠𝑦𝑠𝑡𝑒𝑚
- The mean of the probability distribution of interarrival times is
1/λ = 𝑒𝑥𝑝𝑒𝑐𝑡𝑒𝑑 𝑖𝑛𝑡𝑒𝑟𝑎𝑟𝑟𝑖𝑣𝑎𝑙 𝑡𝑖𝑚𝑒
- Most queueing models assume that the form of the probability distribution of interarrival
times is an exponential distribution

Properties of the Exponential Distribution

- There is a high likelihood of small interarrival times, but a small chance of a very large
interarrival time
- For most queueing systems, the servers have no control over when the customers will
arrive. Randomly
- Having random arrivals means that interarrival times are completely unpredictable, in the
sense that the chance of an arrival in the next minute is always just the same
- The only probability distribution with this property of random arrivals is the exponential
distribution
- Lack of Memory Property: The fact that the probability of an arrival in the next minute is
completely uninfluenced by when the last arrive occurred

The Queue
- Number of customers in the queue: the number of customers waiting for service to begin
 - Number of customers in the system: the number in the queue + the number currently
being served
- Queue capacity: maximum number of customers that can be held in the queue
- Infinite Queue: unlimited number of customers
- FInite Queue: limited number of customers
- Queue Discipline: order in which members of the queue are selected to begin service
1. First come first served
2. Balking (decide not to join)
3. Reneging (leave due to long waiting time)
4. Jockeying (customers switch queues for shorter waiting time)

Service
- When a customer enters service, the elapsed time from the beginning to the end of the
service is referred to as the service time
- Basic queueing models assume that the service time has a particular probability
distribution
- The symbol used for the mean of the service time distribution is 1/u (expected service
time)
- u means mean service rate

Labels for Queueing Models

>>> - / - / -
1. Distribution of interarrival times
2. Distribution of service times
3. Number of servers
- M = exponential distribution
- D = Degenerate distribution
- Gl = general independent interarrival-time distribution (any distribution)
- G = general service-time distribution (any arbitrary distribution)