Modeling changing dependency structure in multivariate time series

We show how to apply the efficient Bayesian changepoint detection techniques of Fearnhead in the multivariate setting. We model the joint density of vector-valued observations using undirected Gaussian graphical models, whose structure we estimate. We show how we can exactly compute the MAP segmentation, as well as how to draw perfect samples from the posterior over segmentations, simultaneously accounting for uncertainty about the number and location of changepoints, as well as uncertainty about the covariance structure. We illustrate the technique by applying it to financial data and to bee tracking data.

[1]  A. Dawid.  Some matrix-variate distribution theory: Notational considerations and a Bayesian application , 1981 .

[2]  J. Hartigan,et al.  Product Partition Models for Change Point Problems , 1992 .

[3]  J. Hartigan,et al.  A Bayesian Analysis for Change Point Problems , 1993 .

[4]  A. Dawid,et al.  Hyper Markov Laws in the Statistical Analysis of Decomposable Graphical Models , 1993 .

[5]  David Heckerman,et al.  Learning Gaussian Networks , 1994, UAI.

[6]  Steffen L. Lauritzen,et al.  Graphical models in R , 1996 .

[7]  Michael I. Jordan.  Graphical Models , 1998 .

[8]  P. Green,et al.  Decomposable graphical Gaussian model determination , 1999 .

[9]  Michael I. Jordan,et al.  Probabilistic Networks and Expert Systems , 1999 .

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

[11]  S. L. Scott.  Bayesian Methods for Hidden Markov Models , 2002 .

[12]  Christophe Andrieu,et al.  Bayesian curve fitting using MCMC with applications to signal segmentation , 2002, IEEE Trans. Signal Process..

[13]  A. Roverato.  Hyper Inverse Wishart Distribution for Non-decomposable Graphs and its Application to Bayesian Inference for Gaussian Graphical Models , 2002 .

[14]  Pat Langley,et al.  Editorial: On Machine Learning , 1986, Machine Learning.

[15]  Refik Soyer,et al.  Bayesian Methods for Nonlinear Classification and Regression , 2004, Technometrics.

[16]  A. Gelman.  Prior distributions for variance parameters in hierarchical models (comment on article by Browne and Draper) , 2004 .

[17]  Paul Fearnhead,et al.  Exact Bayesian curve fitting and signal segmentation , 2005, IEEE Transactions on Signal Processing.

[18]  N. Hengartner,et al.  Structural learning with time‐varying components: tracking the cross‐section of financial time series , 2005 .

[19]  K. Strimmer,et al.  Statistical Applications in Genetics and Molecular Biology A Shrinkage Approach to Large-Scale Covariance Matrix Estimation and Implications for Functional Genomics , 2011 .

[20]  Paul Fearnhead,et al.  Exact and efficient Bayesian inference for multiple changepoint problems , 2006, Stat. Comput..

[21]  P. Fearnhead,et al.  Efficient Online Inference for Multiple Changepoint Problems , 2006, 2006 IEEE Nonlinear Statistical Signal Processing Workshop.

[22]  N. Meinshausen,et al.  High-dimensional graphs and variable selection with the Lasso , 2006, math/0608017.

[23]  James M. Rehg,et al.  Parameterized Duration Mmodeling for Switching Linear Dynamic Systems , 2006, 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06).

[24]  Alexandre d'Aspremont,et al.  Convex optimization techniques for fitting sparse Gaussian graphical models , 2006, ICML.

[25]  P. Fearnhead,et al.  On‐line inference for multiple changepoint problems , 2007 .

[26]  Michael A. West,et al.  Dynamic matrix-variate graphical models , 2007 .

[27]  Gm Gero Walter,et al.  Bayesian linear regression , 2009 .

[28]  D. B. Dahl.  Modal clustering in a class of product partition models , 2009 .