| Tekijä: | Lovejoy, W. S. |
| Otsikko: | Computationally feasible bounds for partially observed Markov decision processes. |
| Lehti: | Operations Research
1991 : JAN-FEB, VOL. 39:1, p. 162-175 |
| Asiasana: | DECISION MODELS MARKOV CHAINS |
| Kieli: | eng |
| Tiivistelmä: | A partially observed Markov decision process /POMDP/ is a sequential decision problem where information concerning parameters of interest is incomplete, and possible actions include sampling, surveying, or otherwise collecting additional information. An approximate solution procedure for POMDP has been constructed based upon a fixed grid approximation to the exact, uncountable infinite space. Using this procedure, an approximate strategy can be generated, and rigorous bounds on the value loss relative to optimal for using that strategy can be calculated. A numerical example illustrates the methodology. |
SCIMA