A Partially Observable Markov Decision Process with Lagged Information
Soung Hie Kim, Byung Ho Jeong · Journal of the Operational Research Society · 1987
In actuality, we face lots of uncertainty in the application of a Markov process. In order to reduce such uncertainty, it is indispensable to use additional information concerning the state of the process.Among various kinds of additional information, this paper focuses on how to use uncertain delayed observation in a partially observable Markovian decision process (POMDP). This study develops a basic information structure, adding lagged observations to a general POMDP, and derives a rule for updating the state vector based on the information structure. This POMDP model is solved on the basis of a modified one-pass algorithm. An example is also given.