https://github.com/aftersol/basic_genetic_algorithm

Basic genetic algorithm that computes tries to find best global maximum of two variables functions
https://github.com/aftersol/basic_genetic_algorithm

Last synced: 7 days ago
JSON representation

Basic genetic algorithm that computes tries to find best global maximum of two variables functions

• Host: GitHub
• URL: https://github.com/aftersol/basic_genetic_algorithm
• Owner: Aftersol
• Created: 2024-08-08T06:46:58.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2024-08-08T21:35:40.000Z (over 1 year ago)
• Last Synced: 2024-08-09T08:31:55.457Z (over 1 year ago)
• Language: Jupyter Notebook
• Homepage:
• Size: 321 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Basic Genetic Algorithm From Scratch

  ![alt text](intro.png "Using elite selection, the best chromosomes over many generations slowly rises starting from 0.60 to 0.67")

This code calculates the best fitness of the equation below:

$$\left(f(x,y)=(1-x)^2*exp(-x^2-(y+1)^2)-(x-x^3-y^3)*exp((-x^2-y^2))\right)$$

Where

$$-2<=x,y<=2$$

## How to run

A Jupyter Notebook software such as Google Colab is required. Copy genetic_ai.ipynb to your Jupyter Notebook software and then run the code.

Elite Selection is enabled by default. You can disable this by setting

  elite_selection=False

Change this variable to set the amount generations the genetic algorithm will run for

  generation=200

You can adjust the fitness equation in the genetic_ai.ipynb provided to find the best fitness of different functions given the variables x and y

Adjust the left and right limits to adjust the range of the decoding function

Increasing the range will decrease the accuracy of the solution

  left_limit = -2.0

  right_limit = 2.0

Adjust the crossover and mutation probability to change how fast each member mutates and crossover with each other within the next generation.

  crossover_prob = 0.7 # probabity of crossover

  mutation_prob = 0.01 # probabity of mutations

## Requirements

- A Jupyter Notebook software such as Google Colab

- Numpy

- Deepcopy