Can I read Evolving Neural Networks Through Augmenting Topologies: Kenneth O. Stanley and Risto Miikkulainen on EtoBox?
Evolving Neural Networks Through Augmenting Topologies: Kenneth O. Stanley and Risto Miikkulainen by Kashinath is a document available to read on EtoBox.
What is Evolving Neural Networks Through Augmenting Topologies: Kenneth O. Stanley and Risto Miikkulainen about?
This paper presents a new method called NeuroEvolution of Augmenting Topologies (NEAT) that evolves both the connection weights and topology of neural networks. NEAT is shown to outperform other methods on a reinforcement learning benchmark by incrementally growing complexity and protecting innovative topologies. The paper claims NEAT is more efficient because it employs principled crossover of topologies, protects innovations, and grows from minimal structures.
- Author
- Kashinath
- Language
- EN