Thursday, 27 September 2012

Evolving neurone networks of optimal complexity

Evolving neurone networks to deal with prisoners dilemma and snowdrift games by McNally, Brown, and Jackson. This is a clever paper that provides one possible route to the social brain hypothesis. They show that if more processing power leads to an advantage in the game then it will emerge. This in itself would be fairly trivial, but they produce this result via agent level evolution of neurone-network-like processing mechanism. Here is the link:

No comments:

Post a Comment