It’s interesting to investigate combination of deep neuro-evolution and self-replication to evolve Artificial Neural Networks (ANNs) able to keep and complexify innate learned structures aimed to fulfill auxiliary tasks (orthogonal to the self-replication).
This project provides GOLang implementation of Neuro-Evolution of Augmented Topologies (NEAT) algorithm which uses Novelty Search optimization to find a solution for deceptive tasks with strong local optima.
This project provides GOLang implementation of Neuro-Evolution of Augmented Topologies (NEAT) algorithm. The neuroevolution methods of ANN training allows us to start with a very simple synthetic organism and evolve it to produce a unit of intelligence that represents an approximation of a complex real-world concept. The training accomplished by gradual complexification of the topology of neural networks that are encoded into the genome of a synthetic intelligence unit.
The neuroevolution methods of ANN training allows us to start with a very simple synthetic organism and evolve it to produce a unit of intelligence that represents an approximation of a complex real-world concept. The training accomplished by gradual complexification of the topology of neural networks that are encoded into the genome of a synthetic intelligence unit. There can be several ANNs joined into the complex hierarchy of modules.