Maximal mean/standard deviation ratio in an undiscounted MDP

DOI10.1016/0167-6377(85)90022-7zbMath0576.90096OpenAlexW1980175358MaRDI QIDQ1064974

Matthew J. Sobel

Publication date: 1985

Published in: Operations Research Letters (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1016/0167-6377(85)90022-7

zbMATH Keywords

average undiscounted reward-ratemaximization of the mean/standard deviation ratiomean/varianceparametric analysis of a linear programstationary probability distribution of the reward

Mathematics Subject Classification ID

Discrete-time Markov processes on general state spaces (60J05) Dynamic programming (90C39) Markov and semi-Markov decision processes (90C40) Applications of Markov renewal processes (reliability, queueing networks, etc.) (60K20)

Related Items (13)

Markov Decision Problems Where Means Bound Variances ⋮ Analysis of a class of fractional programming problems ⋮ Survey of linear programming for standard and nonstandard Markovian control problems. Part I: Theory ⋮ Variance minimization for constrained discounted continuous-time MDPs with exponentially distributed stopping times ⋮ A Fourth bibliography of fractional programming ⋮ On the total reward variance for continuous-time Markov reward chains ⋮ A Solution Strategy for a Class of Nonlinear Knapsack Problems ⋮ A note on maximal mean/standard deviation ratio in an undiscounted MDP ⋮ An Inequality for Variances of the Discounted Rewards ⋮ Constrained Multiagent Markov Decision Processes: a Taxonomy of Problems and Algorithms ⋮ On mean-standard deviation ratio problems and beyond via c-programming ⋮ On mean reward variance in semi-Markov processes ⋮ Notes on average Markov decision processes with a minimum-variance criterion

Cites Work

• Vector-Valued Dynamic Programming
• Constrained Undiscounted Stochastic Dynamic Programming
• Discounted MDP’s: Distribution Functions and Exponential Utility Maximization
• Risk-sensitive linear/quadratic/gaussian control
• Mean-Variance Tradeoffs in an Undiscounted MDP
• Technical Note—Dynamic Programming and Probabilistic Constraints

This page was built for publication: Maximal mean/standard deviation ratio in an undiscounted MDP