Title :
A learning algorithm for Markov decision processes with adaptive state aggregation
Author :
Baras, J.S. ; Borkar, V.S.
Author_Institution :
Inst. for Syst. Res., Maryland Univ., College Park, MD, USA
Abstract :
We propose a simulation-based algorithm for learning good policies for a Markov decision process with unknown transition law, with aggregated states. The state aggregation itself can be adapted on a slower time scale by an auxiliary learning algorithm. Rigorous justifications are provided for both algorithms
Keywords :
Markov processes; adaptive systems; decision theory; learning (artificial intelligence); stochastic systems; Markov decision processes; adaptive state aggregation; learning algorithm; state aggregation; unknown transition law; Algorithm design and analysis; Clustering algorithms; Communication system control; Computational modeling; Computer science; Data compression; Educational institutions; Learning; State estimation; Stochastic processes;
Conference_Titel :
Decision and Control, 2000. Proceedings of the 39th IEEE Conference on
Conference_Location :
Sydney, NSW
Print_ISBN :
0-7803-6638-7
DOI :
10.1109/CDC.2000.912220
