Data-Driven MCMC for Learning and Inference in Switching Linear Dynamic Systems

Show full item record

Please use this identifier to cite or link to this item: http://hdl.handle.net/1853/38601

Title: Data-Driven MCMC for Learning and Inference in Switching Linear Dynamic Systems
Author: Oh, Sang Min ; Rehg, James M. ; Balch, Tucker ; Dellaert, Frank
Abstract: Switching Linear Dynamic System (SLDS) models are a popular technique for modeling complex nonlinear dynamic systems. An SLDS has significantly more descriptive power than an HMM, but inference in SLDS models is computationally intractable. This paper describes a novel inference algorithm for SLDS models based on the Data- Driven MCMC paradigm. We describe a new proposal distribution which substantially increases the convergence speed. Comparisons to standard deterministic approximation methods demonstrate the improved accuracy of our new approach. We apply our approach to the problem of learning an SLDS model of the bee dance. Honeybees communicate the location and distance to food sources through a dance that takes place within the hive. We learn SLDS model parameters from tracking data which is automatically extracted from video. We then demonstrate the ability to successfully segment novel bee dances into their constituent parts, effectively decoding the dance of the bees.
Description: ©2005. American Association for Artificial Intelligence. The original publication is available at: www.aaai.org
Type: Post-print
Proceedings
URI: http://hdl.handle.net/1853/38601
Citation: Oh, S.M., Rehg, J.M., Balch, T., & Dellaert, F. (2005). Data-Driven MCMC for Learning and Inference in Switching Linear Dynamic Systems. Proceedings of the National Conference on Artificial Intelligence (AAAI 2005), 9-13 July 2005, 944-949.
Date: 2005-07
Contributor: Georgia Institute of Technology. Center for Robotics and Intelligent Machines
Georgia Institute of Technology. College of Computing
Publisher: Georgia Institute of Technology
AAAI Press
Subject: Convergence speed
Data-driven MCMC
Datasets
Honeybee dance
Switching linear dynamic system

All materials in SMARTech are protected under U.S. Copyright Law and all rights are reserved, unless otherwise specifically indicated on or in the materials.

Files in this item

Files Size Format View
Oh05aaai.pdf 532.4Kb PDF View/ Open

This item appears in the following Collection(s)

Show full item record