Goal of this program was to train a neural network model through genetic algorithm in gym environment API provided by OpenAI. Main goal of this task force to design an efficient crossover method and mutation method. Then we had to show the fitness graph over generations.
Python
Supported in both Windows and Linux
Not tried yet on Mac
- gym
- NumPy
- matplotlib
- random
Only instruction to run this file is if the graph is not optimal, stuck in local maxima, that is does not achieve a good maximum (say 180) then you can run this code again and that should solve the problem. This is due to the entire random structure of the algorithm overall.