neat4jsmall
Home
Tutorials
Documentation
Screenshots
Downloads
Contact

NEAT4J Description

NEAT4J is a Java framework that implements the NEAT algorithm as proposed by Kenneth O Stanley.

NEAT4J allows you to create your own experiments very easily allowing you to train the population and then use the best inidividual in your application.

NEAT4J not only provides a single threaded local learning environment, but also provides a distributed learning environment allowing remote machines to run experiments, reporting results back to a centralised server.

To view the documentation, learn how to use and extend NEAT4J or download NEAT4J, use the links on the left hand side.


NEAT - A One Liner

NEAT(NeuroEvolution for Augmenting Topologies) is an evolutionary algorithm which, using genetic algorithms, simultaneously evolves both the neuron connection weights and the topology of the NEAT neural network.