Train a Neural Network to play Snake using a Genetic Algorithm
Role in this project:
ML Engineer Contributions:56 commits, 2 PRs, 45 pushes in 4 months
Contributions summary:Greer primarily focused on developing and improving a neural network for playing the Snake game using a genetic algorithm. They implemented the core logic for the Snake AI, including vision calculations, decision-making based on the neural network output, fitness calculations, and genetic algorithm components like crossover and mutation. Further work involved adding features such as saving and loading trained models, a mutation rate adjustment, and a visualization of the evolution process through a graph.
genetic-algorithmneural-network
A Tensorflow implementation of a Conditional GAN for generating human faces from a text description.
Contributions:37 commits, 36 pushes, 1 branch in 4 months
tensorflow