Decision Science Notes
Analytical tools for complex business problems.
Chapter-Wise Module Notes & PDF Downloads
Short Notes
All Modules IncludedComplete syllabus revision notes covering all modules in one single PDF
Module 1
Introduction to Operations Research & Linear Programming
Module 2
The Simplex Method & Duality
Module 3
Transportation Problem Models
Module 4
Assignment Problem Models
Read Module Notes Online
Module I: Foundations of Decision Science & Linear Programming Modeling
Decision Science, historically designated as Operations Research (OR), represents the systematic application of advanced mathematical modeling, statistical analysis, and algorithmic optimization to resolve complex managerial problems. Originating during World War II to optimize military logistics and radar deployment, decision science today provides the quantitative backbone for supply chain planning, airline fleet scheduling, production sequencing, and corporate capital budgeting.
1.1 Nature, Scope, and Methodology of Decision Science
Decision science approaches managerial challenges through a structured scientific methodology consisting of six phases:
- Problem Identification and Formulation: Defining operational objectives, identifying key variables, and establishing system boundaries.
- Mathematical Model Construction: Translating verbal operational relationships into symbolic equations representing an objective function and structural constraints.
- Model Derivation / Solution: Executing mathematical algorithms (e.g., Simplex algorithm, Hungarian method) to identify the optimal solution.
- Model Validation: Testing the model against historical data to ensure behavioral stability and accuracy.
- Sensitivity Analysis: Evaluating how the optimal solution shifts when external technological or cost parameters fluctuate.
- Implementation and Control: Deploying the decision rules into daily enterprise operational software.
1.2 The General Linear Programming Problem (LPP) Formulation
A Linear Programming Problem (LPP) seeks to maximize or minimize a linear objective function subject to a system of linear equality or inequality constraints and non-negativity restrictions.
1.3 Graphical Solution Method (Two Decision Variables)
When an LPP involves strictly two decision variables (), it can be solved geometrically on a two-dimensional Cartesian plane:
- Plot Constraint Boundaries: Treat each inequality as a strict equality line () and plot the line across the x1 and x2 axes.
- Identify the Feasible Region: Determine the permissible half-plane satisfying the inequality direction ( points toward origin, points away from origin). The intersection of all valid half-planes in the first quadrant () forms the convex Feasible Polygon.
- Extreme Point (Corner Point) Theorem: Mathematical linear programming dictates that the optimal solution of an LPP always lies at one of the extreme corner vertices of the feasible region.
- Evaluate Corner Vertices: Compute the objective value Z at each corner point coordinate; the vertex yielding the maximum (or minimum) value represents the global optimum.
1.4 Special Cases in Linear Programming
| Special LPP Condition | Geometric Diagnostic Indicator | Underlying Operational Cause | Managerial Interpretation |
|---|---|---|---|
| Alternative / Multiple Optima | Objective function line is perfectly parallel to a binding constraint line. | Equal profitability trade-offs between products. | Management possesses flexibility to select multiple product mixes with identical profit. |
| Unbounded Solution | Feasible region is open-ended; objective value Z can increase to . | Missing or improperly formulated resource constraint equations. | Flawed model formulation; infinite profit is impossible in reality. |
| Infeasible Solution | Zero common overlapping area satisfies all constraints simultaneously. | Mutually contradictory, overly restrictive operational mandates. | Management cannot satisfy all targets simultaneously; constraints must be relaxed. |
| Degeneracy | More than two constraint lines intersect at a single corner vertex. | Redundant resource limits. | May cause cycling in Simplex algorithmic iterations. |
1.5 The Simplex Method Algorithm & Sensitivity Analysis Overview
For problems involving three or more decision variables, the graphical method fails, necessitating George Dantzig's algebraic Simplex Method:
- Standard Form Conversion: Inequalities are converted into strict equalities by adding Slack Variables () to constraints (representing unused capacity) or subtracting Surplus Variables () and adding Artificial Variables () to constraints (Big-M or Two-Phase Method).
- Algorithmic Iteration: Starting from an initial basic feasible solution (usually origin, where real variables are zero and slack variables equal resource capacities), the algorithm computes the net evaluation row (). The entering variable is selected via the most positive indicator, and the leaving variable is determined via the minimum positive replacement ratio ().
- Optimality Criterion: Optimality is reached when all indicators in the net evaluation row are non-positive () for maximization problems.
Duality and Sensitivity Analysis: Every primal maximization LPP possesses a symmetric dual minimization LPP. The dual variables (yi) represent the Shadow Prices of resources—the maximum premium management should pay for an additional unit of a constrained resource.
Module II: Transportation Logistics Algorithms & Assignment Optimization
Transportation and Assignment models represent specialized subclasses of linear programming designed to optimize logistics distribution networks and human resource matching. In modern enterprise supply chains, minimizing freight shipping costs between geographically dispersed factories and regional distribution centers, and optimally matching personnel to specific operational tasks, directly governs corporate margins and service delivery times.
2.1 The Transportation Problem Formulation
The standard transportation model considers m manufacturing origins (factories) supplying a homogeneous commodity to n demand destinations (warehouses):
A transportation problem is Balanced if total factory capacity equals total market demand (). If an imbalance exists, a fictitious dummy plant or dummy warehouse with zero shipping costs () is introduced to restore mathematical parity.
2.2 Methods for Finding an Initial Basic Feasible Solution (IBFS)
To initialize the optimization process, three classical heuristics are deployed:
- North-West Corner Rule (NWCR): Strictly mechanical allocation starting at cell in the upper left corner, fulfilling row supply or column demand sequentially. Completely ignores unit shipping costs, frequently yielding an economically poor starting solution.
- Least Cost Method (LCM / Matrix Minimum): Allocates the maximum feasible quantity to the cell possessing the lowest unit transportation cost () across the entire matrix, iteratively eliminating exhausted rows and satisfied columns.
- Vogel's Approximation Method (VAM): Computes penalty costs for each row and column (the arithmetic difference between the lowest and next-lowest unit cost in that line). Allocates the maximum feasible volume to the lowest cost cell in the line exhibiting the highest penalty. VAM accounts for opportunity costs, consistently generating an IBFS remarkably close to the global optimum.
2.3 Optimality Testing: The MODI (Modified Distribution) Method
A basic feasible solution containing exactly independent allocations is tested for optimality via the MODI Method ( method):
- Assign dual row values (ui) and column values (vj) such that for every occupied (allocated) cell:
Setting an arbitrary initial =0 allows solving for all remaining ui and vj.
- Compute the opportunity cost (net evaluation) for every unoccupied cell:
- Optimality Benchmark: If all , the current allocation is globally optimal. If any <0, the solution can be improved by forming a closed loop starting from the most negative cell, shifting allocations, and recalculating.
2.4 The Assignment Problem and the Hungarian Method
The Assignment Problem optimizes the one-to-one matching of n jobs to n facilities or workers to minimize total completion cost or execution time (). Because standard simplex or transportation techniques are computationally inefficient for this discrete matrix, Harold Kuhn developed the Hungarian Method based on Dénes Kőnig's matrix reduction theorems:
- Row Reduction: Subtract the minimum element in each row from every element in that row.
- Column Reduction: Subtract the minimum element in each column from every element in that column.
- Line Covering Test: Draw the minimum number of horizontal and vertical lines required to cover all zeros in the reduced matrix. If the number of lines equals the matrix dimension n, an optimal assignment is achievable among the zeros.
- Matrix Modification: If the line count is less than n, identify the smallest uncovered element (k). Subtract k from all uncovered elements and add k to all elements situated at the intersection of two lines. Repeat the line covering test.
- Assignment Execution: Assign jobs to workers starting with rows/columns containing a single zero.
Module III: Project Network Scheduling, CPM Algorithms & PERT Modeling
Project management governs the coordinated execution of non-routine, complex corporate initiatives—such as constructing industrial facilities, implementing ERP software architectures, or executing pharmaceutical clinical trials. Network analysis techniques decompose multi-stage projects into discrete activities and sequential precedence relationships, enabling managers to determine the minimum completion duration, identify critical path bottlenecks, and optimize resource allocations under uncertainty.
3.1 Project Network Architecture and Construction Rules
A project network represents workflow logic utilizing directed graphs:
- Activity: A distinct task requiring time and resources, represented by an arrow (→).
- Event (Node): The starting or ending milestone of an activity, represented by a circle (○). Events consume zero time and zero resources.
- Precedence Relationships: Immediate predecessor and successor activities governing technological dependencies.
- Dummy Activities: Fictitious activities consuming zero time and zero resources, drawn as dashed arrows (&dashrightarrow;), utilized strictly to maintain correct precedence logic or prevent two activities from sharing identical starting and ending nodes.
3.2 Critical Path Method (CPM): Deterministic Scheduling
CPM applies deterministic time estimates where activity durations are known with high certainty. Network traversal involves two computational sweeps:
- Forward Pass (Earliest Times): Computes the Earliest Start Time () and Earliest Finish Time () by moving from project initiation to completion:
- Backward Pass (Latest Times): Computes the Latest Finish Time () and Latest Start Time () by traversing backward from project completion:
The Critical Path is the longest contiguous path of interconnected activities connecting project start to finish. Activities on this path have zero total float (). Any delay in a critical activity directly expands overall project completion time.
3.3 Activity Float (Slack) Metrics
| Float Metric | Mathematical Formula | Managerial Operational Definition |
|---|---|---|
| Total Float () | Maximum duration an activity can be delayed without delaying overall project completion. | |
| Free Float () | Delay allowable without delaying the earliest start time of any immediate successor activity. | |
| Independent Float () | Delay allowable under worst-case scheduling conditions without affecting any other activity. |
3.4 Program Evaluation and Review Technique (PERT): Probabilistic Modeling
Developed for the U.S. Navy Polaris missile project, PERT incorporates time uncertainty by utilizing three distinct duration estimates following a Beta Distribution:
- Optimistic Time (a): Minimum possible duration assuming optimal operational conditions.
- Most Likely Time (m): Modal duration occurring under standard, normal working conditions.
- Pessimistic Time (b): Maximum duration assuming adverse delays and complications.
To evaluate the probability of completing the project within a target scheduled deadline (Ts), the standard normal variable Z is calculated:
The resulting Z-score maps directly to standard cumulative normal probability tables, delivering defensible project risk metrics to executive leadership.
Module IV: Game Theory Strategies & Decision-Making Under Uncertainty
Managerial decision-making occurs in environments where strategic competitors actively retaliate against corporate initiatives, or where future states of nature (economic growth, regulatory shifts) cannot be forecasted with certainty. Game theory models competitive interaction between rational, self-interested agents, while classical decision theory provides structured mathematical criteria to evaluate outcomes under risk and radical uncertainty.
4.1 Foundations of Game Theory: Two-Person Zero-Sum Games
In strategic gaming, a Two-Person Zero-Sum Game represents a competitive conflict where the economic gain achieved by Player A exactly equals the economic loss incurred by Player B (). Payoffs are structured in a matrix where positive numbers denote payouts from Player B to Player A:
4.1.1 Pure Strategy Games and Saddle Point Equilibrium
In a pure strategy game, both players adopt a single defensive strategy governed by minimax and maximin logic:
- Maximin Principle (Player A): Player A anticipates worst-case payoffs for each strategy row (identifies row minima) and selects the strategy that maximizes these minimum gains ().
- Minimax Principle (Player B): Player B anticipates worst-case losses for each strategy column (identifies column maxima) and selects the strategy that minimizes these maximum losses ().
If , the game possesses a Saddle Point Equilibrium. The corresponding matrix value is the Value of the Game (V). If a saddle point exists, neither player has any incentive to unilaterally deviate from their equilibrium pure strategy.
4.1.2 Mixed Strategy Games: The Principle of Dominance
When , no saddle point exists. Players must randomize their choices over a probability distribution of strategies to keep competitors guessing. Before solving, matrices are reduced using the Principle of Dominance:
- Row Dominance: If every element in Row i is less than or equal to the corresponding element in Row k, Row i is strictly dominated and can be deleted (Player A will never choose an inferior payoff).
- Column Dominance: If every element in Column j is greater than or equal to the corresponding element in Column m, Column j is strictly dominated and can be deleted (Player B seeks to minimize payouts).
For reduced matrices, optimal probability weights ( and ) and the game value are computed analytically via the oddment method:
4.2 Decision Theory Under Uncertainty and Risk
Decision situations are categorized by the degree of informational certainty: Decision Making Under Certainty (deterministic), Decision Making Under Risk (probabilistic states of nature), and Decision Making Under Uncertainty (probabilities completely unknown).
4.2.1 Decision Criteria Under Complete Uncertainty
When objective or subjective probabilities cannot be assigned to future states of nature, managers deploy structured behavioral heuristics across a payoff matrix:
| Decision Criterion | Managerial Psychological Stance | Mathematical Selection Rule |
|---|---|---|
| MaxiMax | Extreme Optimism (Risk-seeking gambler). | Identify maximum payoff for each act; select the maximum of the maximums (). |
| MaxiMin (Wald's Rule) | Extreme Pessimism / Conservatism. | Identify minimum payoff for each act; select the maximum of the minimums (). |
| Minimax Regret (Savage) | Opportunity Loss Minimization. | Construct regret matrix (); select act minimizing maximum regret. |
| Laplace (Equal Likelihood) | Principle of Insufficient Reason. | Assign equal probability () to all states of nature; select act with highest expected payoff. |
| Hurwicz (Criterion of Realism) | Balanced realism via coefficient of optimism (). | Select act maximizing: . |
4.2.2 Decision Making Under Risk: EMV and EVPI
When probabilities (pj) can be assigned to states of nature, decisions are governed by the Expected Monetary Value (EMV) criterion:
The Expected Value of Perfect Information (EVPI) represents the maximum price management should pay to obtain perfect predictive intelligence before committing capital:
Calicut University FYUGP Academic Framework
Syllabus details for BBA Semester 4 under official Calicut University regulations.
Assessment comprises a 70-mark End-Semester Examination (2 Hours) and 30-mark Continuous Internal Evaluation.
About Decision Science (Calicut University BBA Sem 4)
Decision Science (BBA4CJ205 • 4 Credits • 60 Hours) is the central quantitative operations research core paper in Calicut University BBA Semester 4 under the FYUGP curriculum. Designed to build mathematical problem-solving and optimization capabilities for future business leaders, the course provides rigorous instruction in deterministic and stochastic operations research models used to resolve complex resource allocation, transportation, assignment, project scheduling, and queuing challenges. The curriculum is structured around classic mathematical optimization models. Undergraduates begin with Linear Programming Problems (LPP), learning to translate verbal corporate resource constraints into mathematical objective functions and linear constraints, solving two-variable problems via the graphical method, and mastering the multi-iteration Simplex algorithm (slack, surplus, artificial variables, Big-M method, and duality principles). The course delivers comprehensive training in Transportation Problems, teaching initial basic feasible solution methods (North-West Corner Rule, Least Cost Method, Vogel's Approximation Method - VAM) and the Modified Distribution (MODI) method to test and achieve optimal distribution cost. Students master the Hungarian Assignment Method for optimal one-to-one task allocation, including unbalanced, prohibited, and maximization variations. Furthermore, the course provides exhaustive training in Project Management Network Techniques, contrasting the deterministic Critical Path Method (CPM) with the probabilistic Program Evaluation and Review Technique (PERT) three-time estimates, critical path identification, float calculations, project variance, and completion probability estimation.
Decision Science Module-by-Module Curriculum Breakdown
Official syllabus unit modules prescribed by the Calicut University FYUGP framework:
Introduction to Operations Research & Linear Programming
Establishes foundational concepts of operations research and mathematical modeling. Topics include the historical development, definition, scope, and phases of operations research in business decision-making; classification of OR models; Linear Programming Problem (LPP) formulation: defining decision variables, constructing the linear objective function (maximization or minimization), and formulating linear constraint equations and non-negativity restrictions; graphical solution method: plotting constraint boundary lines, identifying the feasible region (bounded vs unbounded), corner point/extreme point evaluation method; and graphical interpretations of unique optimal solutions, multiple optimal solutions, infeasible solutions, and unbounded solutions.
The Simplex Method & Duality
Delivers rigorous mathematical training in the algebraic Simplex optimization algorithm. Topics include converting LPP into standard canonical form; introduction of slack variables (for <= constraints), surplus variables, and artificial variables (for >= and = constraints); setting up the initial Simplex tableau; determining the pivot column (most negative Cj - Zj value for maximization), pivot row (minimum positive ratio rule), and key/pivot element; executing Gaussian row transformations to update the tableau; optimality condition (all Cj - Zj <= 0); handling artificial variables using the Big-M Method (Charnes' Penalty Method); and Duality in LPP: formulation of the Dual from the Primal problem, economic interpretation of dual variables (shadow prices), and fundamental duality theorems.
Transportation Problem Models
Examines mathematical distribution models to minimize freight transportation costs between multiple sources and destinations. Topics include mathematical formulation of the transportation problem; balanced versus unbalanced transportation problems (introducing dummy source/destination); methods for finding Initial Basic Feasible Solution (IBFS): North-West Corner Rule (NWCR), Least Cost Method (LCM / Matrix Minimum Method), and Vogel's Approximation Method (VAM - penalty calculation method); testing for degeneracy in transportation problems (m + n - 1 allocated cells condition); and testing for optimality using the Modified Distribution Method (MODI method / u-v method), calculating net evaluations (dij = cij - (ui + vj)) for unoccupied cells, and tracing closed loops to redistribute allocations and achieve minimum total transportation cost.
Assignment Problem Models
Covers one-to-one resource allocation optimization to minimize cost or maximize productivity. Topics include mathematical formulation of the assignment problem as a special case of the transportation model; the Hungarian Assignment Method (Flood's technique): row reduction step, column reduction step, drawing minimum number of horizontal and vertical lines to cover all zeros, modifying the matrix, and making final optimal zero assignments; variations in assignment problems: unbalanced assignment problems (adding dummy rows or columns with zero costs), prohibited/restricted assignments (assigning high penalty cost M to impossible routes), and maximization assignment problems (converting to minimization by subtracting all matrix elements from the highest element).
Network Analysis & Queuing Theory
Addresses project scheduling optimization and waiting line operational analysis. Project Network Analysis covers fundamentals of network construction (nodes, arrows, events, activities: concurrent, predecessor, successor, dummy activities); rules for drawing valid network diagrams (avoiding looping and dangling); Critical Path Method (CPM): forward pass calculation (Earliest Event Time - TE, Earliest Start Time - ES, Earliest Finish Time - EF), backward pass calculation (Latest Event Time - TL, Latest Start Time - LS, Latest Finish Time - LF), calculation of Total Float, Free Float, and Independent Float, and identifying the Critical Path; Program Evaluation and Review Technique (PERT): stochastic three-time estimates (Optimistic time 'a', Most likely time 'm', Pessimistic time 'b'), expected activity time formula (te = (a + 4m + b)/6), activity variance (sigma^2 = ((b - a)/6)^2), project standard deviation, and calculating probability of project completion by a specified deadline using standard normal Z-tables; and Queuing Theory basics: components of a queuing system (arrival process, service mechanism, queue discipline), and single-server queuing model (M/M/1: infinity/FIFO) formulas for server utilization, average queue length, and waiting time.
Core Curriculum & Key Topics
- Foundations of Operations Research: modeling phases, assumptions, and managerial applications
- Linear Programming Problem (LPP): mathematical formulation of objective functions and constraints
- Graphical method for LPP: feasible region identification, corner-point method, and special cases (unbounded, infeasible, multiple optima)
- Transportation problem: Initial Basic Feasible Solution (IBFS) via North-West Corner Rule (NWCR), Least Cost Method (LCM), and Vogel's Approximation Method (VAM)
- Optimality testing in transportation: Modified Distribution (MODI) method and stepping-stone method
- Assignment Problem: Hungarian Assignment Method for minimization and unbalanced/maximization variations
- Network analysis: Critical Path Method (CPM), Program Evaluation and Review Technique (PERT), three time estimates (to, tm, tp), and critical path identification
- Game Theory: two-person zero-sum games, saddle point, dominance principles, and graphical solution of 2xn / mx2 games
Learning Outcomes
- Formulate real-world corporate resource constraints into Linear Programming mathematical models.
- Solve transportation distribution problems to minimize multi-warehouse freight costs using VAM and MODI methods.
- Apply the Hungarian Method to assign jobs to machines or executives minimizing total execution time.
- Calculate project completion timelines and critical paths using CPM and PERT network scheduling techniques.
Calicut University Examination Scheme for Decision Science
The evaluation framework for Decision Science (BBA4CJ205) under Calicut University FYUGP regulations combines Continuous Internal Evaluation (30 Marks) with the university End-Semester Examination (70 Marks) for a total of 100 Marks (2 Hours duration):
10 Questions × 3 Marks (Maximum ceiling 24 Marks). Short definitional and conceptual questions covering all 4 syllabus modules.
8 Questions × 6 Marks (Maximum ceiling 36 Marks). Paragraph questions, comparative evaluations, and analytical problems.
2 Questions × 10 Marks (Answer any 1 for 10 Marks). Long-form essay or comprehensive numerical workout problem.
Exam Preparation Strategy:
To achieve maximum marks in Calicut University Decision Science (BBA4CJ205) examinations: • In Part A: State operations research definitions and mathematical conditions precisely. For Transportation Degeneracy, state that degeneracy occurs when the number of allocated cells is less than (m + n - 1), where m is the number of origins and n is the number of destinations. For PERT expected time, write te = (a + 4m + b) / 6. • In Part B: In Transportation Problems, always calculate the initial solution using Vogel's Approximation Method (VAM) because it yields the lowest initial cost. Clearly show your row and column penalty calculations outside the matrix table. In Assignment problems, show row reduction and column reduction steps in separate clearly labeled matrices. • In Part C (Comprehensive Mathematical Workouts): Simplex method, MODI optimality testing, and CPM/PERT network diagrams are frequent 15-mark questions. For Network Analysis, draw large, neat network diagrams using a ruler. Label each node with its event number, Earliest Time (TE), and Latest Time (TL). Construct a complete network analysis table with columns: Activity, Duration, ES, EF, LS, LF, Total Float, Free Float, and Critical status (Yes/No). Clearly state the Critical Path (e.g. 1-2-4-6) and Total Project Duration in bold. Allocate 20 minutes to Part A, 45 minutes to Part B, and 85 minutes to Part C workouts.
Exam Preparation & Study Guides — Decision Science
Why is Vogel's Approximation Method (VAM) preferred over the North-West Corner Rule for transportation problems?
NWCR simply allocates based on geographic grid position without considering unit shipping costs, often yielding high initial costs. VAM computes unit cost penalty differences for each row and column, allocating to lowest-cost routes first, yielding an Initial Basic Feasible Solution much closer to mathematical optimality.
How do CPM and PERT differ in project management network analysis?
CPM is a deterministic model used for repetitive, predictable projects (like construction) where task durations are known with certainty. PERT is a probabilistic model used for novel R&D projects with uncertain durations, relying on three probabilistic time estimates (optimistic, most likely, pessimistic).
What is a Saddle Point in Game Theory?
A saddle point is a game equilibrium where the Maximin value of the row player equals the Minimax value of the column player. In such games, both players possess a pure strategy that maximizes their minimum payoff, eliminating any incentive to alter choices.
What are the core mathematical assumptions of Linear Programming Models (LPM)?
Linear programming assumes Proportionality (the contribution of each variable to the objective function and constraints is directly proportional), Additivity (the total effect of variables is the exact sum of individual effects), Divisibility (fractional values for decision variables are permissible), and Certainty (all objective and constraint parameters are known constants).
How does Vogel's Approximation Method (VAM) determine an initial basic feasible solution in transportation problems?
VAM computes penalty costs for each row and column representing the difference between the lowest and next-lowest unit costs, then allocates maximum possible cargo units to the lowest-cost cell in the row or column bearing the largest penalty, generally yielding near-optimal starting solutions.
What is the Critical Path in PERT/CPM network analysis and why is its float zero?
The Critical Path is the longest sequential chain of connected activities through a project network diagram, dictating the minimum total calendar time required to complete the project. Activities on this path have zero total float (slack); any delay in a critical activity directly postpones the overall project completion date.
Course outlines and study notes follow official Calicut University curricula. Curated by Hadi (Kozhikode).
For feedback or corrections, contact support@degreelive.in.
Related Study Resources & Student Support
Related Semester 4 Subjects
Browse study notes for other courses in BBA Sem 4:
Financial Aid & Grants 2026-27
Check eligibility criteria and deadlines for Kerala government and merit scholarships (e-Grantz 3.0, KSHEC, DCE).
Request Notes
Can't find what you're looking for? Let us know which subject notes you need.