0502902 20240903170642.120190314235959.9 85064190063 000457070200009 10.14736/kyb-2018-6-1218 Risk-sensitive Average Optimality in Markov Decision Processes 13 s. P cav_un_epca*02971630023-5954Kybernetika546 (2018)1218-1230Ústav teorie informace a automatizace AV ČR, v. v. i. controlled Markov processes risk-sensitive average optimality asymptotic behavior cav_un_auth*0101196 Department of Econometrics 100% Sladký Karel UTIA-B Ekonometrie Department of Econometrics E E K Ústav teorie informace a automatizace AV ČR, v. v. i. http://library.utia.cas.cz/separaty/2019/E/sladky-0502902.pdf cav_un_auth*0363963 GA18-02739S GA ČR In this note attention is focused on finding policies optimizing risk-sensitive optimality criteria in Markov decision chains. To this end we assume that the total reward generated by the Markov process is evaluated by an exponential utility function with a given risk-sensitive coefficient. The ratio of the first two moments depends on the value of the risk-sensitive coefficient, if the risk-sensitive coefficient is equal to zero we speak on risk-neutral models. Observe that the first moment of the generated reward corresponds to the expectation of the total reward and the second central moment of the reward variance. For communicating Markov processes and for some specific classes of unichain processes long run risk-sensitive average reward is independent of the starting state. In this note we present necessary and sufficient condition for existence of optimal policies independent of the starting state in unichain models and characterize the class of average risk-sensitive optimal policies. cav_un_auth*0373581 Mathematical Methods in Economy and Industry 2017 20170904 20170906 Jindřichův Hradec CZ WOS BB 10000 10100 10103 2019 1 RVO:67985556 http://hdl.handle.net/11104/0295273 S 1 Article|Proceedings Paper Computer Science Cybernetics COMPUTERSCIENCE.CYBERNETICS 0.591 0.155 13 0.00068 0.174 891 0.268 0.500 71 Q4 15.991 6.5 Q4 Q3 0.17 Q4 6.522 2018 85064190063 SCOPUS PUBMED 000457070200009 WOS cav_un_epca*0297163 Kybernetika 0023-5954 Roč. 54 č. 6 2018 1218 1230 Ústav teorie informace a automatizace AV ČR, v. v. i.