flâneur

NEAT Overview — NEAT-Python 1.1.0 documentation

neat-python.readthedocs.io · 534 words · saved by 1 readers

NEAT (NeuroEvolution of Augmenting Topologies) is an evolutionary algorithm that creates artificial neural networks. For a detailed description of the algorithm, you should probably go read some of Stanley’s papers on his website. Even if you just want to get the gist of the algorithm, reading at least a couple of the early NEAT papers is a good idea. Most of them are pretty short, and do a good job of explaining concepts (or at least pointing you to other references that will). The initial NEAT paper is only 6 pages long, and Section II should be enough if you just want a high-level overview. In the current implementation of NEAT-Python, a population of individual genomes is maintained. Each genome contains two sets of genes that describe how to build an artificial neural network: Node genes, each of which specifies a single neuron. Connection genes, each of which specifies a single connection between neurons. To evolve a solution to a problem, the user must provide a fitness function

NEAT (NeuroEvolution of Augmenting Topologies) is an evolutionary algorithm that creates artificial neural networks. For a detailed description of the algorithm, you should probably go read some of Stanley’s papers on his website. Even if you just want to get the gist of the algorithm, reading at least a couple of the early NEAT papers is a good idea. Most of them are pretty short, and do a good job of explaining concepts (or at least pointing you to other references that will). The initial NEAT paper is only 6 pages long, and Section II should be enough if you just want a high-level…

saved by

related reading