Overview This project is an implementation of an AI that learns to play the popular game Flappy Bird using the NeuroEvolution of Augmenting Topologies (NEAT) algorithm. The AI is capable of adapting its gameplay strategy over time through evolutionary neural networks, demonstrating the power of machine learning and evolutionary computation.
Features AI-Powered Gameplay: The AI uses NEAT to evolve neural networks that control the Flappy Bird character, learning and optimizing its actions to survive longer in the game. Real-Time Learning: As the AI plays, it continually improves its performance by evolving better strategies through genetic algorithms. Performance Visualization: Track the progress and performance of the AI with visual feedback, including metrics like survival time and score. Customizable Parameters: Tweak NEAT parameters such as population size, mutation rates, and fitness functions to experiment with different AI behaviors and outcomes. Installation Prerequisites Python 3.6+ pygame for running the Flappy Bird game environment neat-python for implementing the NEAT algorithm Setup Clone the repository: git clone https://github.com/nomodify/flappy-bird-ai-neat.git cd flappy-bird-ai-neat Install dependencies: pip install -r requirements.txt Run the AI: python flappy_bird_ai.py
Certainly! Below is a sample README for your Flappy Bird AI project using NEAT in Python, tailored for a GitHub repository:
Flappy Bird AI using NEAT (Python) Overview This project is an implementation of an AI that learns to play the popular game Flappy Bird using the NeuroEvolution of Augmenting Topologies (NEAT) algorithm. The AI is capable of adapting its gameplay strategy over time through evolutionary neural networks, demonstrating the power of machine learning and evolutionary computation.
Features AI-Powered Gameplay: The AI uses NEAT to evolve neural networks that control the Flappy Bird character, learning and optimizing its actions to survive longer in the game. Real-Time Learning: As the AI plays, it continually improves its performance by evolving better strategies through genetic algorithms. Performance Visualization: Track the progress and performance of the AI with visual feedback, including metrics like survival time and score. Customizable Parameters: Tweak NEAT parameters such as population size, mutation rates, and fitness functions to experiment with different AI behaviors and outcomes. Installation Prerequisites Python 3.6+ pygame for running the Flappy Bird game environment neat-python for implementing the NEAT algorithm Setup Clone the repository:
bash Copy code git clone https://github.com/yourusername/flappy-bird-ai-neat.git cd flappy-bird-ai-neat Install dependencies:
bash Copy code pip install -r requirements.txt Run the AI:
bash Copy code python flappy_bird_ai.py How It Works NEAT Algorithm NEAT (NeuroEvolution of Augmenting Topologies) is an evolutionary algorithm that creates and evolves neural networks. In this project, NEAT is used to evolve a population of neural networks, where each network represents a potential Flappy Bird player. The AI’s fitness is determined by how long the bird survives in the game, with the best-performing networks being selected and mutated to create the next generation.
Game Environment The AI interacts with a clone of the Flappy Bird game built using pygame. The game provides inputs such as the bird’s position and the distance to the next pipe, which the neural network processes to decide whether the bird should flap its wings.
Evolutionary Process Initialization: The NEAT algorithm starts with a random population of neural networks. Evaluation: Each network is tested in the game environment, and its fitness score is calculated based on survival time and score. Selection and Mutation: The best networks are selected to reproduce, introducing mutations and crossover to create the next generation. Iteration: This process repeats for many generations, with each generation improving over the last. Customization The AI's behavior can be adjusted by modifying the NEAT configuration file (config-feedforward.txt). Key parameters include:
Population Size: Number of neural networks in each generation. Mutation Rates: Probability of mutations in the network structure. Fitness Function: Criteria for evaluating the success of each network. Experimenting with these parameters can lead to different AI behaviors and learning efficiencies.
Results
Over time, the AI learns to survive longer in the game, demonstrating the effectiveness of NEAT in training neural networks for complex tasks. The progress of the AI can be visualized through graphs and performance metrics displayed during runtime.
