CSE Publications - Report Abstract
CSE-2004-17
Title : Automated Hierarchical POMDP Construction through Data-mining Techniques
Type : Technical Report
Author(s) : G. Michael Youngblood, Edwin O. Heierman, Diane J. Cook, and Lawrence B. Holder
Abstract : Markov models provide a useful representation of system behavioral actions and state observations, but they do not scale well. Utilizing a hierarchy and abstraction as in HHMMs improves scalability, but they are usually constructed manually using knowledge engineering techniques. In this paper, we introduce a new method of automatically constructing HHMMs and subsequent hierarchical POMDPs using the output of a sequential data-mining algorithm. We present the theory of this technique and frame it in a case study involving a learning living room.