Sharing Features among Dynamical Systems with Beta Processes

We propose a Bayesian nonparametric approach to the problem of modeling related time series. Using a beta process prior, our approach is based on the discovery of a set of latent dynamical behaviors that are shared among multiple time series. The size of the set and the sharing pattern are both inferred from data. We develop an efficient Markov chain Monte Carlo inference method that is based on the Indian buffet process representation of the predictive distribution of the beta process. In particular, our approach uses the sum-product algorithm to efficiently compute Metropolis-Hastings acceptance probabilities, and explores new dynamical behaviors via birth/death proposals. We validate our sampling algorithm using several synthetic datasets, and also demonstrate promising results on unsupervised segmentation of visual motion capture data.

[1]  J. Kingman,et al.  Completely random measures. , 1967 .

[2]  Lawrence R. Rabiner,et al.  A tutorial on hidden Markov models and selected applications in speech recognition , 1989, Proc. IEEE.

[3]  N. Hjort.  Nonparametric Bayes Estimators Based on Beta Processes in Models for Life History Data , 1990 .

[4]  C. Hwang,et al.  Convergence rates of the Gibbs sampler, the Metropolis algorithm and other single-site updating dynamics , 1993 .

[5]  P. Green.  Reversible jump Markov chain Monte Carlo computation and Bayesian model determination , 1995 .

[6]  Rakesh Dugad,et al.  A Tutorial On Hidden Markov Models , 1996 .

[7]  Jun S. Liu.  Peskun's theorem and a modified discrete-state Gibbs sampler , 1996 .

[8]  Michael A. West,et al.  Bayesian forecasting and dynamic models (2nd ed.) , 1997 .

[9]  Vladimir Pavlovic,et al.  A dynamic Bayesian network approach to figure tracking using learned dynamic models , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[10]  Vladimir Pavlovic,et al.  Learning Switching Linear Models of Human Motion , 2000, NIPS.

[11]  Carl E. Rasmussen,et al.  Factorial Hidden Markov Models , 1997 .

[12]  David S. Touretzky,et al.  Model Uncertainty in Classical Conditioning , 2003, NIPS.

[13]  Jernej Barbic,et al.  Segmenting Motion Capture Data into Distinct Behaviors , 2004, Graphics Interface.

[14]  Kari Pulli,et al.  Style translation for human motion , 2005, SIGGRAPH 2005.

[15]  Thomas L. Griffiths,et al.  Infinite latent feature models and the Indian buffet process , 2005, NIPS.

[16]  Zoubin Ghahramani,et al.  Modeling Dyadic Data with Binary Latent Factors , 2006, NIPS.

[17]  Geoffrey E. Hinton,et al.  Modeling Human Motion Using Binary Latent Variables , 2006, NIPS.

[18]  Carl E. Rasmussen,et al.  A choice model with infinitely many latent features , 2006, ICML.

[19]  Michael I. Jordan,et al.  Hierarchical Dirichlet Processes , 2006 .

[20]  Thomas L. Griffiths,et al.  A Non-Parametric Bayesian Method for Inferring Hidden Causes , 2006, UAI.

[21]  B. Schölkopf,et al.  Modeling Human Motion Using Binary Latent Variables , 2007 .

[22]  Michael I. Jordan,et al.  Hierarchical Beta Processes and the Indian Buffet Process , 2007, AISTATS.

[23]  Michael I. Jordan,et al.  An HDP-HMM for systems with state persistence , 2008, ICML '08.

[24]  David J. Fleet,et al.  This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE Gaussian Process Dynamical Model , 2007 .

[25]  Yee Whye Teh,et al.  The Infinite Factorial Hidden Markov Model , 2008, NIPS.

[26]  Michael I. Jordan,et al.  Nonparametric Bayesian Learning of Switching Linear Dynamical Systems , 2008, NIPS.