A Hybrid Genetic/Optimization Algorithm for Finite-Horizon, Partially Observed Markov Decision Processes

  title={A Hybrid Genetic/Optimization Algorithm for Finite-Horizon, Partially Observed Markov Decision Processes},
  author={Zong-Zhi Lin and J. Bean and C. White},
  journal={INFORMS J. Comput.},
  • Zong-Zhi Lin, J. Bean, C. White
  • Published 2004
  • Mathematics, Computer Science
  • INFORMS J. Comput.
  • The partially observed Markov decision process (POMDP) is a generalization of a Markov decision process that allows for noise-corrupted and costly observations of the underlying system state. The value function of the infinite horizon POMDP is known to be piecewise affine and convex in the probability mass vector over the state space. Such a function can be represented by a finite set of affine functions.In this paper, we develop and evaluate an exact algorithm, GAMIP, which combines a genetic… CONTINUE READING
    Partially Observable Markov Decision Processes
    • 69
    • PDF
    Reinforcement Learning: A Tutorial Survey and Recent Advances
    • 225
    • PDF
    Evolutionary policy iteration for solving Markov decision processes
    • 43
    • PDF
    A survey on metaheuristics for stochastic combinatorial optimization
    • 469
    • Highly Influenced
    • PDF
    Using evolution strategies to solve DEC-POMDP problems
    • 11
    Value Function Discovery in Markov Decision Processes With Evolutionary Algorithms
    • 11
    • Highly Influenced
    • PDF
    Combining (Integer) Linear Programming Techniques and Metaheuristics for Combinatorial Optimization
    • 101
    • PDF


    Publications referenced by this paper.
    A survey of algorithmic methods for partially observed Markov decision processes
    • 574
    The Optimal Control of Partially Observable Markov Processes over a Finite Horizon
    • 1,499
    Markov Decision Processes: Discrete Stochastic Dynamic Programming
    • 10,197
    • PDF
    Algorithms for partially observable markov decision processes
    • 126
    • Highly Influential
    Exact and approximate algorithms for partially observable markov decision processes
    • 413
    Acting Optimally in Partially Observable Stochastic Domains
    • 661
    • PDF
    Planning in Stochastic Domains: Problem Characteristics and Approximation
    • 68