![]() ![]() The theory of reinforcement learning provides a normative account, deeply rooted in psychological and neuroscientific perspectives on animal behaviour, of how agents may optimize their control of an environment. Machine Learning Algorithm which we have adopted in this project is based on the concept of Neuro-Evolution and this form of machine learning uses algorithms which can evolve and mature over time such as a genetic algorithm to train artificial neural networks. Phaser Framework was used to facilitate HTML5 programming for introducing real-life factors like gravity, collision and Synaptic Neural Network library was used to implement neural network so as to avoid creating a neural network from scratch. The gameplay was divided into two level of difficulty to facilitate study on the learning process. We first create an agent which learns how to optimally play the famous “Flappy Bird” game by safely dodging all the barriers and flapping its way through them and then study the effect of changing various parameters like number of neurons on the hidden layer, gravity, speed, gap between trees has on the learning process. The aim of this paper is to develop and study an artificial intelligence based game-playing agent using genetic algorithm and neural networks. ![]()
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |