Tasks for Evolutionary Computing Course at ITMO University
The goal is to play around basic implementation of Genetic Algorithm and tune it in a way that it will able to find approximate optimal solution of given function. The target function is modification of Rastrgin function. It is reversed, displaced and scaled in such a way that result values are in range [0.0, 10.0]. Possible range of input variables are [-5, 5] for all variables.
Deap framework was used.
It was required that the algorithm must find solutions with fitness values around 9.9, and the algorithm successfully found solutions with best fitness value equal to 9.954.
The goal is to develop an Evolutionary algorithm to solve the queens puzzle. This is classic optimisation problem, where you have to allocate N queens across NxN chessboard in such a way, so they will not conflict to each other. The N is a parameter and can be varied. Try to think how to represent your solutions and how to perform mutation and crossover.
Deap framework was used.
The solution is based on the following tutorial.