Wojciech Mojsiejuk | Genetic Algorithm

Photo by Alexander Popov on Unsplash

This is my final project for Artificial Intelligence course from 4th semester (Summer Semester 2018/19), conducted in cooperation with Pawel.

For function parsing we used SymPy symbols and lambdify function which results then were computed using NumPy ndarrays.

PyQt5 as a GUI, Matplotlib as a data visualisation.

Features

Defining function of two variables, its domain and range

Parsing function

Function visualization in 3D with possibility to manipulate angles of a 3D object

Defining crossover, mutation probability, number of individuals and epochs in GA

Plot of each generation

Summary with chromosome which values will give the best solution

Plot of best, median and mean value in each epoch

Screenshots

Functions visualization, setting function domain and range

\[x^2 + y^2; x \in \langle -2, 2 \rangle y \in \langle -2, 2 \rangle\]

\[x^3 + y^2; x \in \langle -3, 3 \rangle y \in \langle -4, 4 \rangle\]

\[-x^2 + 3 \cdot y; x \in \langle -3, 3 \rangle y \in \langle -4, 4 \rangle\]

Solving genetic algorithm

Generation plots

Heatmap and points representing chromosomes' value. First generation (left), last generation (right).

Summary

Chromosome values which will result in the best solution. Plot for best, median and mean in each generation.

Remaster

In 2019 Pawel and I decided to update this project further and with our colleague Daniel we created a new version for Building Desktop Applications using WPF course from 6th Semester (Summer Semester 2019/20). We rewrite whole project to C# and add additional featues such as: