论文信息 - The Recurrent Control Neural Network

The Recurrent Control Neural Network

This paper presents our Recurrent Control Neural Network (RCNN), which is a model-based approach for a data-efficient modelling and control of reinforcement learning problems in discrete time. Its architecture is based on a recurrent neural network (RNN), which is extended by an additional control network. The latter has the particular task to learn the optimal policy. This method has the advantage that by using neural networks we can easily deal with high-dimensions or continuous state and action spaces. Furthermore we can profit from their high systemidentification and approximation quality. We show that our RCNN is able to learn a potentially optimal policy by testing it on two different settings of the mountain car problem.

[1] R. Sutton,et al. Theoretical Results on Reinforcement Learning with Temporally Abstract Behaviors , 1998 .

[2] Doina Precup,et al. Theoretical Results on Reinforcement Learning with Temporally Abstract Options , 1998, ECML.

[3] Geoffrey E. Hinton,et al. Learning internal representations by error propagation , 1986 .

[4] Jürgen Schmidhuber,et al. Reinforcement Learning in Markovian and Non-Markovian Environments , 1990, NIPS.

[5] Pieter Bram Bakker,et al. The state of mind : reinforcement learning with recurrent neural networks , 2004 .

[6] H. Wechsler,et al. Competitive reinforcement learning in continuous control tasks , 2003, Proceedings of the International Joint Conference on Neural Networks, 2003..

[7] Andrew W. Moore,et al. Direct Policy Search using Paired Statistical Tests , 2001, ICML.

[8] S. Udluft,et al. A Recurrent Control Neural Network for Data Efficient Reinforcement Learning , 2007, 2007 IEEE International Symposium on Approximate Dynamic Programming and Reinforcement Learning.

[9] Michail G. Lagoudakis,et al. Least-Squares Policy Iteration , 2003, J. Mach. Learn. Res..