Scribd
Upload
3 views3 pages

Genetic Algorithm AI Tutorials 2025

Genetic Algorithms (GAs) are optimization techniques inspired by natural selection, involving steps such as initialization, selection, crossover, mutation, and termination. The process begins with a randomly generated population of candidate solutions, followed by selecting parents based on fitness, combining them to create offspring, and introducing mutations to maintain diversity. Ultimately, GAs aim to evolve solutions to reach optimal or near-optimal results through iterative cycles.

Uploaded by

kabyoroy3
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
3 views3 pages

Genetic Algorithm AI Tutorials 2025

Genetic Algorithms (GAs) are optimization techniques inspired by natural selection, involving steps such as initialization, selection, crossover, mutation, and termination. The process begins with a randomly generated population of candidate solutions, followed by selecting parents based on fitness, combining them to create offspring, and introducing mutations to maintain diversity. Ultimately, GAs aim to evolve solutions to reach optimal or near-optimal results through iterative cycles.

Uploaded by

kabyoroy3
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Artificial Intelligence 101

Rajdeep Chatterjee, Ph.D.


Amygdala AI, Bhubaneswar, India *
January 2024

Genetic Algorithms

P
1 Introduction to Genetic Algorithms

EE
Genetic Algorithms (GAs) are search heuristics inspired by the process of natural selection. They are used to
solve optimization and search problems by evolving solutions over generations. The steps involved in a GA
are:
1. Initialization
JD
2. Selection
3. Crossover
A

4. Mutation
5. Termination
ER

2 Step-by-Step Process
2.1 Initialization
A population of candidate solutions (chromosomes) is generated randomly. Each solution is represented as a
CS

binary string.
Example: Consider a toy population for minimizing the function f (x) = x2 .

Chromosome Decimal Value (x) Fitness ( f (x) = x2 )


1010 10 100
0100 4 16
1110 14 196
0011 3 9
* Amygdala AI, is an international volunteer-run research group that advocates for AI for a better tomorrow [Link]
org/.

1
2

2.2 Selection
Select parents based on their fitness. A common method is roulette wheel selection, where solutions with
lower fitness are more likely to be chosen.
Example: Assign probabilities based on fitness:
Chromosome Fitness Probability Cumulative Probability
1010 100 0.32 0.32
0100 16 0.52 0.84
1110 196 0.11 0.95
0011 9 0.05 1.00

2.3 Crossover
Combine pairs of parents to create offspring. A single-point crossover is used, where a random point in the
chromosome is selected, and the genes are swapped beyond that point.

P
Example: Crossover between 1010 and 0100 at position 2:

Parent 1: 1010

EE
Parent 2: 0100
Offspring 1: 1000
Offspring 2: 0110
JD
2.4 Mutation
Introduce small random changes to maintain diversity and avoid premature convergence.
Example: Mutate a bit in 1000 (flip the 2nd bit):
A

Original: 1000
Mutated: 1100
ER

2.5 Termination
The algorithm terminates when a stopping criterion is met, such as a maximum number of generations or
convergence to an optimal solution.
CS

3 Visual Representation
4 Conclusion
The genetic algorithm evolves a solution by mimicking natural selection. Through repeated cycles of selec-
tion, crossover, and mutation, the algorithm converges to an optimal or near-optimal solution for the given
problem.

Copyright ©cserajdeep, 2024


3

Initialize Population

P
Evaluate Fitness

Select Parents
EE Check Termination
JD
Perform Crossover
A
ER

Apply Mutation
CS

New Population

Copyright ©cserajdeep, 2024

You might also like