Skip to content

Reinforcement Learning (NEAT) integrated with a popular one button game - Flappy Bird

Notifications You must be signed in to change notification settings

Grace-Hephzibah/Flappy-Bird

Repository files navigation

Flappy-Bird

Reinforcement Learning (NEAT) integrated with a popular one-button game - Flappy Bird

Objective

The objective of this project is to develop a Flappy Bird game using the Pygame library and implement the NEAT (NeuroEvolution of Augmenting Topologies) algorithm to train an agent capable of mastering the Flappy Bird game. Customized NEAT parameters will be employed to optimize the training process, and upon successful training, the best-performing model will be serialized using the pickle format for future use and analysis. Through this project, we aim to demonstrate the effectiveness of the NEAT algorithm in evolving intelligent behaviours and producing a Flappy Bird-playing agent that achieves high levels of performance.

SnapShot