Evolving neural networks

摘要

Neuroevolution, i.e. evolution of artificial neural networks, has recently emerged as a powerful technique for solving challenging reinforcement learning problems. Compared to traditional (e.g. value-function based) methods, neuroevolution is especially strong in domains where the state of the world is not fully known: The state can be disambiguated through recurrency, and novel situations handled through pattern matching. In this tutorial, I will review (1) neuroevolution methods that evolve fixed-topology networks, network topologies, and network construction processes, (2) ways of combining traditional neural network learning algorithms with evolutionary methods, and (3) applications of neuroevolution to control, robotics, artificial life, and games.