ESANN
2008
14 years 1 months ago
2008
Natural policy gradient methods and the covariance matrix adaptation evolution strategy, two variable metric methods proposed for solving reinforcement learning tasks, are contrast...