Go back to UGC NET COMUPTER SCIENCE
UGC NET Computer Science Application – OPTIMIZATION
(Linear Programming · Simplex · Dual Simplex · Sensitivity · Integer Programming · Transportation · Assignment · PERT · CPM · Resource Levelling · Cost Considerations)
1. LINEAR PROGRAMMING (LP)
Definition: Linear Programming
Linear programming is a mathematical technique for optimizing (maximizing or minimizing) a linear objective function, subject to a set of linear constraints.
General structure:
-
Objective function
-
Subject to constraints:
-
with non-negativity constraints:
Explanation
LP is used where resources (labor, materials, machines) are limited and must be allocated optimally.
Simple Example
Maximize
Subject to:
2. MATHEMATICAL MODEL (FORMULATION)
Definition
Formulation is the process of expressing a real-world optimization problem as an LPP using:
-
Decision variables
-
Objective function
-
Constraints
Explanation
This converts a word problem into equations.
Example
A factory makes chairs (x) and tables (y).
Profit = 50x + 80y
Wood constraint → 3x + 5y ≤ 30
Labor constraint → 2x + y ≤ 10
3. GRAPHICAL SOLUTION METHOD
Definition
A method to solve LPPs with two variables by drawing constraints on a 2D graph and identifying the feasible region.
Explanation
-
Convert inequalities into lines.
-
Shade feasible region.
-
Compute objective value at each vertex.
-
Best value gives the optimum.
Example
Line x + y = 4 divides the plane; feasible region lies below it.
4. SIMPLEX METHOD
Definition
The Simplex Method is an iterative algebraic method to solve LPPs with two or more variables.
Explanation
-
Converts inequalities to equalities by adding slack, surplus, artificial variables
-
Builds a tableau
-
Performs pivoting to improve the solution
-
Stops when no negative indicators remain in the objective row
Key Terms
-
Basic variables: variables with non-zero values
-
Non-basic variables: set to zero
-
Pivot element: used for row operations
5. DUAL SIMPLEX METHOD
Definition
A modification of simplex used when:
-
The objective row is feasible
-
The basic solution is infeasible (negative RHS)
Explanation
Dual Simplex works by making the RHS feasible through pivoting until all RHS ≥ 0.
6. SENSITIVITY ANALYSIS
Definition
Sensitivity analysis studies how changes in:
-
Coefficients of objective function
-
RHS of constraints
-
Constraint coefficients
affect the optimal solution.
Key Terms
-
Shadow price: change in optimal Z when RHS of a constraint increases by 1 unit.
-
Reduced cost: amount by which objective coefficient must improve before variable enters basis.
-
Allowable range: how much parameters can vary without altering the optimal solution.
7. INTEGER PROGRAMMING (IP)
Definition
An optimization model where some or all decision variables must be integers.
Types
-
Pure Integer Programming – all variables integer
-
Mixed Integer Programming – some integer
-
0–1 (Binary) Programming – variables can be 0 or 1
Explanation
Used where fractional decisions are impossible (e.g., number of machines can’t be 2.5).
Methods
-
Branch & Bound
-
Cutting Plane
-
Dynamic Programming
8. TRANSPORTATION PROBLEM
Definition
A special LPP that deals with transporting commodities from multiple sources to multiple destinations at minimum cost.
Explanation
-
Rows: sources
-
Columns: destinations
-
Each cell: cost of transporting one unit
Balance condition:
Total supply = Total demand
Initial Solution Methods
-
North-West Corner
-
Least Cost Method
-
Vogel’s Approximation Method (VAM)
Optimal Solution Methods
-
MODI (u – v method)
-
Stepping-Stone Method
9. ASSIGNMENT PROBLEM
Definition
A special case of transportation where:
-
Supply = demand = 1
-
One person does one task only
Goal: Minimize cost or maximize efficiency.
Explanation
Square cost matrix (NxN).
Solved by Hungarian Method.
Steps
-
Row reduction
-
Column reduction
-
Covering zeros
-
Optimal assignment
10. PERT (PROGRAM EVALUATION REVIEW TECHNIQUE)
Definition
PERT is a project management tool using probabilistic time estimates to compute expected project duration.
Explanation
Each activity has three time estimates:
-
= optimistic time
-
= most likely time
-
= pessimistic time
Expected Time
Variance
Used when time is uncertain.
11. CPM (CRITICAL PATH METHOD)
Definition
A deterministic technique to determine the longest duration path (critical path) through a project network.
Explanation
-
Activities have certain times
-
Critical Path = longest path
-
Activities on this path have zero float
Float Types
-
Total Float (TF)
-
Free Float (FF)
-
Independent Float (IF)
Critical Path Formula
Critical path = sequence of activities with
12. CRITICAL PATH CALCULATIONS
Steps
1. Forward Pass (Earliest Times)
-
Compute Earliest Start (ES)
-
Compute Earliest Finish (EF = ES + duration)
2. Backward Pass (Latest Times)
-
Compute Latest Finish (LF)
-
Compute Latest Start (LS = LF – duration)
3. Slack Calculation
Activities with Slack = 0 → critical.
13. RESOURCE LEVELLING
Definition
Resource levelling adjusts activity start times to smooth out resource usage over time.
Explanation
Prevents:
-
sudden peaks in manpower
-
idle resources
-
overburdening specific days
Used for optimizing:
-
labor
-
machines
-
budgets
14. COST CONSIDERATION IN PROJECT SCHEDULING
Definition
Scheduling must consider both direct and indirect costs to obtain the lowest total project cost.
Direct Costs
Cost of performing activity itself.
Indirect Costs
Administrative, supervision, overhead.
Crash Cost
Additional cost for reducing activity duration.
Explanation
Project duration vs cost curve is U-shaped.
Goal → Find minimum total cost.
COMPLETE CHAPTER SUMMARY (Exam Sheet)
| Concept | Definition | Keyword |
|---|---|---|
| LP |
Optimize linear objective under linear constraints |
linear + feasible |
| Simplex | Table-based root-finding | pivot |
| Dual Simplex | Fix infeasible RHS |
RHS negative |
| Sensitivity | Post-optimal impact study |
shadow price |
|
Integer Programming |
Variables restricted to integers | branch & bound |
| Transportation |
Min cost supply–demand problem |
VAM, MODI |
| Assignment | One-to-one allocation | Hungarian |
| PERT | Probabilistic time |
expected time |
| CPM | Deterministic time |
critical path |
| Resource Levelling | Smooth resource usage | no peaks |
| Cost Scheduling | Total = Direct + Indirect | crashing |
