"3-540-40506-2" . "Optim\u00E1ln\u00ED \u0159\u00EDzen\u00ED nediskontovan\u00FDch markovsk\u00FDch rozhodovac\u00EDch procesu s penalizac\u00ED rozptylem"@cs . . . . "2002-03-11+01:00"^^ . "Laxenburg" . "Optimal solutions for undiscounted variance penalized Markov decision chains"@en . . "Markov decision processes;optimal policies;mean-variance"@en . "RIV/67985556:_____/04:00106191!RIV/2005/GA0/A16005/N" . . . . "Sita\u0159, M." . "RIV/67985556:_____/04:00106191" . "578165" . "1"^^ . . "Optimal solutions for undiscounted variance penalized Markov decision chains"@en . "Optimal solutions for undiscounted variance penalized Markov decision chains" . . "[A515C5691CEA]" . "P(GA402/01/0539), P(GA402/02/1015), Z(AV0Z1075907)" . "Dynamic Stochastic Optimization" . "We investigate how the minimum variance criterion can work in discrete stochastic dynamic programmig. We adapt notions and notation used in Markov decision processes and in contrast to the classical models we also considered variance of the obtained total reward. Various mean-variance optimality criteria are discussed and an algorithmical procedure for finding optimal policies is suggested. An illustrative numerical example is supplied."@en . . . "Optim\u00E1ln\u00ED \u0159\u00EDzen\u00ED nediskontovan\u00FDch markovsk\u00FDch rozhodovac\u00EDch procesu s penalizac\u00ED rozptylem"@cs . "2"^^ . "43;66" . "Optimal solutions for undiscounted variance penalized Markov decision chains" . "V pr\u00E1ci se studuj\u00ED vlastnosti krit\u00E9ri\u00ED optimality pro \u00FAlohy diskr\u00E9tn\u00EDho stochastick\u00E9ho dynamick\u00E9ho programov\u00E1n\u00ED zahrnuj\u00EDc\u00ED t\u00E9\u017E rozptyl celkov\u00E9ho v\u00FDnosu. Jsou vy\u0161et\u0159ov\u00E1ny r\u016Fzn\u00E9 varianty krit\u00E9ri\u00ED optimality typu pr\u016Fm\u011Brn\u00FD v\u00FDnos a jeho rozptyl a jsou navr\u017Eeny algoritmick\u00E9 postupy pro nalezen\u00ED optim\u00E1ln\u00EDch \u0159\u00EDzen\u00ED vzhledem k uva\u017Eovan\u00FDm krit\u00E9ri\u00EDm. Pr\u00E1ce je dopln\u011Bna ilustrativn\u00EDm numerick\u00FDm p\u0159\u00EDkladem"@cs . . . . "We investigate how the minimum variance criterion can work in discrete stochastic dynamic programmig. We adapt notions and notation used in Markov decision processes and in contrast to the classical models we also considered variance of the obtained total reward. Various mean-variance optimality criteria are discussed and an algorithmical procedure for finding optimal policies is suggested. An illustrative numerical example is supplied." . . "24"^^ . . "Springer-Verlag" . "Sladk\u00FD, Karel" . . . . "Berlin" .