Objective: Find values of variables \( x_1,

x_2, ..., x_n \) that maximize or minimize an

objective function within given constraints.

Traveling Salesman Problem (TSP)

Optimization Problem Overview
Knapsack Problem

Examples of Optimization Problems: Graph Coloring Problem

Job Assignment Problem

Binary Search Tree Construction

1. Unconstrained: No specific boundaries

on the variables.

2. Constrained: Constraints or limitations

are applied to variable values.

3. Integer Programming: Variables are

Single-Objective Optimization: Use GAs to restricted to integer values.
find the best solution for problems like
TSP, job scheduling, or graph coloring. Types of Optimization Problems
4. Real-valued: Variables can take any real
number within specified limits.
Multi-Objective Optimization: GAs can be Solving Optimization Problems
Introduction to Optimization
adapted for problems with multiple Problems 5. Mixed-integer: Some variables are
conflicting objectives by optimizing each integers, while others are real numbers.
objective separately and combining them.

6. Linear and Non-linear Optimization:

Traveling Salesman Problem (TSP): Objective functions are linear or non-linear.
Minimize the total distance in a path that
visits each city exactly once. Applications of Genetic Minimum Spanning Tree
Algorithms
Knapsack Problem: Maximize the value of
Shortest Path
items in a knapsack without exceeding
weight limits.
Discrete Optimization Problems Examples: Maximum Flow
Case Studies and Applications
Graph Coloring: Assign colors to graph
vertices so that no adjacent vertices share Transportation and Network Design
the same color. Problems

Job Assignment and Scheduling: Assign 1. High computational cost.

tasks to resources optimally.
Limitations of Traditional Optimization 2. Incompatibility with parallel computing.
Approaches
3. Difficulty with discrete variables.
1. Encoding: Represents a solution in a
suitable format for the algorithm.
4. Limited adaptability.
2. Convergence: Determines when the
algorithm should stop.
Introduction: First developed by Prof. John
3. Mating Pool: Set of individuals selected Holland in 1965, with a foundational article
to produce offspring. published in 1975.

4. Fitness Evaluation: Assesses solution Genetics (Gregor Mendel): Study of

Background of Genetic Algorithms (GA)
quality. Key Operations heredity and gene functions.

5. Crossover: Combines parent solutions Principles: Evolution (Charles Darwin): Process of

to create new offspring. natural selection, where the fittest
organisms survive and reproduce.
6. Mutation: Introduces new genes to
maintain diversity. 1. DNA: Spiral helix structure carrying the
genetic code, unique to each species.
7. Inversion: Allows movement from one
optimum to another. 2. Genes: Basic hereditary units, found
Genetic Algorithm Operators within chromosomes, that influence traits.
1. Initial Population: Randomly generate a Basic Concepts of Genetics and Evolution
starting population. 3. Chromosomes: Thread-like structures in
Genetic Algorithms - Concept of Genetics and cells containing genes.
2. Fitness Evaluation: Evaluate each Evolution
solution in the population. I Humans: 46 chromosomes
4. Species Examples:
3. Selection: Choose parents based on Frogs: 26 chromosomes
their fitness scores. Detailed Framework for Genetic Algorithms
1. Information Propagation (Heredity):
4. Crossover and Mutation: Produce Offspring inherit characteristics from their
offspring with diversity. parents.

5. Replacement: Select individuals for the 2. Population Diversity: Variations in

next generation. offspring create diversity in the population.

Key Processes in Biological Evolution 3. Survival of the Fittest (Selection): The

best-suited individuals survive and
1. Population: A set of possible solutions. reproduce.

2. Chromosome: A single solution. 4. Mutation: Introduces new genetic

variations, essential for maintaining
3. Gene: A unit in the chromosome. diversity.

4. Allele: The specific value of a gene.

Genetic Algorithm (GA): A population-
5. Genotype: Representation of the based probabilistic search and
solution in computational space. Terminology in Genetic Algorithms optimization method inspired by biological
evolution and natural selection.
6. Phenotype: The actual solution in real-
world terms. Applications: Frequently used to solve
Definition of Genetic Algorithm
complex optimization problems that are
7. Fitness Function: A function that computationally intensive or impossible to
evaluates solution suitability. solve exactly.

8. Genetic Operators: Selection, crossover, Iterative Process: Repeats a series of steps

mutation, etc. over generations.

Population Size (N): Total solutions in the Probabilistic Search: Uses randomization
population. to search for solutions.

Mating Pool Size (Np): Percentage of Different Genetic Algorithm Key Characteristics of Genetic Algorithms Termination Criteria: Can stop based on
population selected for reproduction. Architectures reaching a local or global optimum, or after
Parameters: Basic Genetic Algorithm a set number of generations.
Mutation Rate (μ): Probability of mutation. Framework
Flexible: GAs can handle "noisy" or complex
Convergence Threshold (δ): Condition for environments and optimize multiple
stopping. objectives.
Generational Genetic Algorithm (GGA)
1. Modular: Solutions can be combined with
Overlapping generations; some individuals
other methods.
are replaced each generation.

Characteristics: 2. Robust: Effective in "noisy"

Biased selection favors highly fit
environments.
individuals.
GA Architectures
Advantages of Genetic Algorithms 3. Flexible: Supports hybrid applications.
Smaller generation gap.

4. Inherently Parallel: Can be distributed

Only a few individuals are replaced each
across multiple systems.
generation. Characteristics:

5. Continuously Improving: Produces

Works well for small populations with long
progressively better solutions.
chromosomes. Steady-State Genetic Algorithm (SSGA)

May get stuck in local optima if there is

insufficient diversity. Limitations: