252 search results - page 7 / 51 » Learning Partially Observable Action Models: Efficient Algor... |
106
click to vote
ATAL
2008
Springer
15 years 1 months ago
2008
Springer
Decentralized Markov decision processes are frequently used to model cooperative multi-agent systems. In this paper, we identify a subclass of general DEC-MDPs that features regul...
ECCV
2004
Springer
16 years 1 months ago
2004
Springer
We present a vision based, adaptive, decision theoretic model of human facial displays in interactions. The model is a partially observable Markov decision process, or POMDP. A POM...
111
click to vote
CVPR
2004
IEEE
16 years 1 months ago
2004
IEEE
We present Propagation Networks (P-Nets), a novel approach for representing and recognizing sequential activities that include parallel streams of action. We represent each activi...
IJCAI
2003
15 years 29 days ago
2003
Filtering denotes any method whereby an agent updates its belief state—its knowledge of the state of the world—from a sequence of actions and observations. In logical filterin...
ICML
2004
IEEE
16 years 12 days ago
2004
IEEE
Predictive state representations (PSRs) have recently been proposed as an alternative to partially observable Markov decision processes (POMDPs) for representing the state of a dy...