On Near Optimality of the Set of Finite-State Controllers for Average Cost POMDP

References

• Aberdeen D., Baxter J. Internal-state policy-gradient algorithms for infinite-horizon POMDPs. (2001) . Technical report, RSISE, Australian National University, Canberra, AustraliaGoogle Scholar
• Baxter J., Bartlett P. L. Infinite-horizon policy-gradient estimation. J. Artificial Intelligence Res. (2001) 15:319–350Crossref, Google Scholar
• Bertsekas D. P., Shreve S.Stochastic Optimal Control: The Discrete Time Case (1978) (Academic Press, New York) Google Scholar
• Bierth K. J. An expected average reward criterion. Stochastic Process. Appl. (1987) 26:123–140Crossref, Google Scholar
• Dynkin E. B., Yushkevich A. A.Controlled Markov Processes (1979) (Springer-Verlag, New York) Crossref, Google Scholar
• Feinberg E. A. An ϵ-optimal control of a finite Markov chain with an average reward criterion. Theory Probab. Appl. (1980) 25:70–81Crossref, Google Scholar
• Feinberg E. A. Controlled Markov processes with arbitrary numerical criteria. Theory Probab. Appl. (1982) 27:486–503Crossref, Google Scholar
• Feinberg E. A. Nonrandomized Markov and semi-Markov strategies in dynamic programming. Theory Probab. Appl. (1982) 27:116–126Crossref, Google Scholar
• Fernández-Gaucherand E., Arapostathis A., Marcus S. I. On the average cost optimality equation and the structure of optimal policies for partially observable Markov decision processes. Ann. Oper. Res. (1991) 29:439–470Crossref, Google Scholar
• Hsu S.-P., Chuang D.-M., Arapostathis A. On the existence of stationary optimal policies for partially observed MDPs under the long-run average cost criterion. Systems Control Lett. (2006) 55:165–173Crossref, Google Scholar
• Jaakkola T. S., Singh S. P., Jordan M. I. Reinforcement learning algorithm for partially observable Markov decision problems. Proc. Neural Inform. Processing Systems Conf. (1995) Denver, CO(MIT Press, Cambridge, MA) Google Scholar
• Lauritzen S. L.Graphical Models (1996) (Oxford University Press, Oxford, UK) Crossref, Google Scholar
• Meuleau N., Peshkin L., Kim K.-E., Kaelbling L. P. Learning finite-state controllers for partially observable environment. Proc. 15th Conf. Uncertainty in Artificial Intelligence (1999) Stockholm, Sweden(Morgan Kaufmann, San Francisco) Google Scholar
• Platzman L. K. Optimal infinite-horizon undiscounted control of finite probabilistic systems. SIAM J. Control Optim. (1980) 18(4):362–380Crossref, Google Scholar
• Puterman M. L.Markov Decision Processes: Discrete Stochastic Dynamic Programming (1994) (John Wiley and Sons, Inc., New York) Crossref, Google Scholar
• Ross S. M. Arbitrary state Markovian decision processes. Ann. Math. Statist. (1968) 39(6):2118–2122Crossref, Google Scholar
• Runggaldier W. J., Stettner L.Approximations of Discrete Time Partially Observable Control Problems, Applied Mathematics Monographs (1994) 6(Giardini Editori e Stampatori, Pisa, Italy) Google Scholar
• Yu H. A function approximation approach to estimation of policy gradient for POMDP with structured polices. Proc. 21st Conf. Uncertainty in Artificial Intelligence (2005) Edinburgh, UK(AUAI Press)Google Scholar
• Yu H. Approximate solution methods for partially observable Markov and semi-Markov decision processes. (2006) . Ph.D. thesis, Massachusetts Institute of Technology, Cambridge, MAGoogle Scholar