xviniette/FlappyLearning
FlappyLearning
FlappyLearning ist eine Sandbox für maschinelles Lernen und ein Trainer für neuronale Netze, der darauf ausgelegt ist, neuroevolutionäre Strategien zu simulieren. Es bietet eine Umgebung, in der künstliche Agenten mittels genetischer Algorithmen entwickelt werden, um ihre Leistung innerhalb einer Spielsimulation zu optimieren.
Das System verwendet ein neuroevolutionäres Populationsmodell, das Feedforward-Neuronale-Netze nutzt, um autonome Spielagenten zu entwickeln. Es steuert die Evolution dieser Agenten durch generationsbasierte Iteration und verwendet fitnessbasierte Selektion, um zu bestimmen, welche Netzwerke überleben und sich reproduzieren.
Das Trainings-Framework integriert stochastische Gewichts-Mutationen zur Erkundung neuer Verhaltensweisen sowie Elitismus-basierte Erhaltung, um die leistungsstärksten Merkmale über Generationen hinweg zu bewahren. Benutzer können Parameter wie Populationsgröße und Mutationsraten anpassen, um den Lernprozess zu optimieren.
Features
- Neuroevolution Frameworks - Provides a complete framework for evolving populations of neural networks across multiple generations.
- Game-Playing Agents - Develops autonomous agents that learn to master a video game environment without manual instruction.
- Weight Mutation - Introduces stochastic weight mutations to explore new behaviors and prevent the population from converging prematurely.
- Fitness-Based Selection - Evaluates agents based on game scores to determine which networks are selected for reproduction.
- Generation-Based Evolutions - Cycles through discrete training phases where each population is replaced by a refined set of descendants.
- Population-Based Training - Implements a population-based training strategy to evolve neural network weights through mutation and selection.
- Feedforward Architectures - Uses a feedforward neural network architecture to map game state inputs to binary action outputs.
- Neural Network Trainers - Implements a trainer that optimizes network weights through a population-based evolutionary loop.
- Neuroevolution Training - Trains neural networks using evolutionary algorithms to solve a game by rewarding high scores.
- Population-Based Optimization - Maintains an evolving group of candidate neural networks to optimize game performance.
- AI Game Simulations - Simulates a Flappy Bird game clone as an environment for evolving neural networks.
- Elitism Strategies - Employs elitism-based preservation to ensure the most successful neural networks are carried forward into new generations.
- Machine Learning Sandboxes - Provides a programmatic space for testing neuroevolutionary strategies and observing the learning progress of agents.
- Evolutionary Parameter Tuning - Allows tuning of population sizes and mutation rates to optimize how the model evolves over time.
