Learning of Dynamic Multilayer Neural networks using Parameter Free PSO

スポンサーリンク

概要

• 論文の詳細を見る

• Recently, learning of continuous trajectories for the identification and/or control of dynamic systems is attractingattention in the application of neural networks. In this paper, we focus on a parameter-free Particle Swarm Optimizationcalled TRIBES, as training algorithm of the dynamic multilayer neural networks (DMNNs). DMNN can be approximatedby a class of the recurrent neural networks with one hidden layer to an arbitrary degree of accuracy. The neural networktraining ability of TRIBES is demonstrated though the computer simulations. In the simulations, TRIBES is compared withother PSO algorithms. As a result, it is confirmed that TRIBES is efficient and practical for learning of continuoustrajectories using DMNNs.

• 2009-03-31

著者

• Norio Yoshida/hiroshi

  Shonan Institute Of Technology/shonan Institute Of Technology

• Yoshida Norio

  Shonan Institute of Technology/Shonan Institute of Technology

関連論文

• Learning of Dynamic Multilayer Neural networks using Parameter Free PSO

スポンサーリンク