Title :
Some new results on the ergodic control of partially observed Markov chains
Author :
Chuang, Dong-Ming ; Arapostathis, Ari
Author_Institution :
Dept. of Electr. & Comput. Eng., Texas Univ., Austin, TX, USA
Abstract :
We derive the dynamic programming equations for the ergodic control problem of partially observed Markov chains with finite state space and compact action space under a rather weak condition imposed on the optimal discounted policies
Keywords :
Markov processes; dynamic programming; optimal control; probability; state-space methods; statistical mechanics; compact action space; dynamic programming equations; ergodic control; finite state space; optimal discounted policies; partially observed Markov chains; Costs; Dynamic programming; Equations; History; Kernel; Optimal control; Probability distribution; Quality control; State-space methods; Time measurement;
Conference_Titel :
Decision and Control, 1999. Proceedings of the 38th IEEE Conference on
Conference_Location :
Phoenix, AZ
Print_ISBN :
0-7803-5250-5
DOI :
10.1109/CDC.1999.830914