emredan / geneticalgorithms

2 Different Genetic Algorithms application. One for the Asymmetric Travelling Salesman Problem and the other is Function Minimization with binary values.

Two different Genetic Algorithm Applications

This project consists of 2 different Genetic Algorithms applications. One for the Asymmetric Travelling Salesman Problem and the other is Function Minimization with binary values.

#Asymmetric Travelling Salesman Problem The map for the consists of 10 cities. Purpose of the application is to find the minimum total cost(time). Every population consists of 8 chromosomes with randomly assigned initial values. For each chromosome, the fitness function value is the sum of the costs within a whole route. -> Please check initilize_map function to see default route costs.

Methodologies used in this Genetich Algorithm application: *Tournament Selection method is used to select parents. Tournament size can be changed as long as not bigger than population size. *Order One crossover method is used for Chromosome Crossover. *For mutation, Insert Mutation is used.

Best results were achieved with Mutation rate of 0.2 and Tournament size of 3

#Function Fitness Minimization

The population consists of 6 chromosomes. Each chromosome consists of 8 bits (boolean values) and since it is signed values the variance is -128 <-> 127. The fitness value is result of the function F(x) = x^2 + 16*x

Methodologies used in this Genetich Algorithm application: *Tournament Selection method with Elitism is used to select parents. Tournament size can be changed as long as not bigger than population size. -->Elitism with 2 parent is the default. To change it see function "Replace_Worst_Two" *Single point crossover method is used for Chromosome Crossover. *For mutation, Binary (swap) Mutation is used. With the rate of 0.2