POMDP, Classification and Regression: Relationships and Joint Utilization
Held in conjunction with ICAPS'06
16th International Conference on Automated Planning & Scheduling
June 6-10th, 2006 Ambleside, The English Lake District, U.K
New! The papers and slides have been posted.
The schedule is now available. There is a moderated discussion after the technical presentations. All participants of the workshop are welcome to participate in the discussion and answer the many challenging open questions posted in the schedule. In addition, the participants are welcome to propose their own questions relevant to the workshop theme. If you want your questions posted in advance in the workshop schedule, you can do so by sending your questions to firstname.lastname@example.org.
Partially observable Markov decision process (POMDP) is a popular model for planning under uncertainty. Classification and regression are standard statistical tools for reconstructing a source (or its attributes) from noise-corrupted data. Studies of POMDPs and classification/regression have been mostly pursued independently in the past. Recently, however, there have emerged a number of papers reporting using classification/regression techniques to solve POMDPs or using a POMDP to build cost-sensitive classifiers.
Much work, however, is still underway in exploring the possibilities of how POMDP and classification/regression techniques can be applied to each other in a mutually beneficial way. The aim of this workshop is to bring together researchers from the POMDP community and researchers from the statistical learning community, and to create an opportunity for exchanging views and reporting on-going work on how a POMDP and a classifier/regressor can mutually benefit each other.
The possibilities of research on this subject have not at all been explored to their full extent and it is time to bring this new interdisciplinary area to the attention of additional researchers. We believe that a broader range of contributions will be stimulated to both POMDP and classification/regression by looking at them from new and unified perspectives.
This is a full-day workshop, consisting of invited and contributed presentations and having an emphasis on interactive discussions.Back to top
Kearns et al.  showed that the concept "sample complexity" used in classification can be extended to the POMDP, and they established an upper bound on the number of trajectories that must be used to insure good generalization. Their work is pioneering in trajectory-based methods and in relating POMDP to classification.
Several researchers investigated using modern classifiers like the SVM to learn MDP policies, including Dietterich and Wang , Lagoudakis and Parr , and Blatt and Hero . Bagnell et al.  reported some preliminary results on classification-based policy search in POMDPs, and Langford and Zadrozny  did some theoretic analysis on this. Mahadeva  and Li et al.  studied the regression methods in POMDPs.
Along the contrary line, Dimitrakakis and Bengio  reported using MDP as a gating network in mixture of experts; Bonet and Geffner , Guo  applied POMDP techniques to classification problems in which the class features and mis-classification are cost-sensitive. The main drawback of the methods in [9-10] is that the features are assumed independent. Relaxation of this naive Bayes assumption is studied in  and encouraging results are reported.
The work in [1-12] signals nontrivial relationships between POMDPs and classification/regression that can be utilized to the benefits of both.
We seek submissions of contributed work in answering the many challenging questions that are summarized in the following topics. Submissions on related topics are also welcome.
Authors are encouraged to submit papers electronically in PDF format. Papers must be formatted using the AAAI style template and must not exceed 10 pages in length. Please send submissions by e-mail to either email@example.com or firstname.lastname@example.org.Back to top
Alfred Hero , University of Michigan at Ann Arbor, USA