A sparse sampling algorithm for near-optimal planning in large Markov decision processes

Machine Learning

Volumn 49, Issue 2-3, 2002, Pages 193-208

  Kearns, Michael ^a   Mansour, Yishay ^b   Ng, Andrew Y ^c  

^a UNIVERSITY OF PENNSYLVANIA   (United States)

^b TEL AVIV UNIVERSITY   (Israel)

^c UNIVERSITY OF CALIFORNIA   (United States)

Author keywords

Markov decision processes; Planning; Reinforcement learning

Indexed keywords

APPROXIMATION THEORY; COMPUTATIONAL COMPLEXITY; MARKOV PROCESSES; MATHEMATICAL MODELS; OPTIMIZATION; PLANNING; STATE SPACE METHODS;

MARKOV DECISION PROCESS; REINFORCEMENT LEARNING ALGORITHMS; SPARSE SAMPLING ALGORITHM;

LEARNING ALGORITHMS;

EID: 0036832951     PISSN: 08856125     EISSN: None     Source Type: Journal    
DOI: 10.1023/A:1017932429737     Document Type: Article

References (18)

1. Learning to act using real-time dynamic programming
2. Exploiting structure in policy construction
3. Tractable inference for complex stochastic processes
4. A robust and fast action selection mechanism for planning
5. Integrating planning and execution in stochastic domains
6. Applying online-search to reinforcement learning
7. Approximate planning in large POMDPs via reusable trajectories
8. Real-time heuristic search
9. Computing factored value functions for policies in structured MDPs
10. Solving robot navigation problems with initial pose uncertainty using real-time heuristic search
11. Finite-sample convergence rates for Q-learning and indirect algorithms
12. Solving very large weakly coupled Markov decision processes
13. Approximate planning for factored POMDPs using belief state simplification
14. An upper bound on the loss from approximate optimal-value functions