Title :
Nonparametric Bayesian factor analysis of multiple time series
Author :
Ray, Priyadip ; Carin, Lawrence
Author_Institution :
Dept. of Electr. & Comput. Eng., Duke Univ., Durham, NC, USA
Abstract :
We propose a nonparametric Bayesian factor analysis framework for characterization of multiple time-series. The proposed model automatically infers the number of factors and the noise/residual variance, and it is also able to cluster time series which behave similarly over prescribed time windows. We use a Pitman-Yor process to impose such clustering. We also provide a general MCMC inference scheme and demonstrate the proposed framework on the analysis of multi-year stock prices of companies in the S & P 500.
Keywords :
Bayes methods; Markov processes; Monte Carlo methods; inference mechanisms; pattern clustering; pricing; stock markets; time series; Monte Carlo Markov chain inference scheme; Pitman-Yor process; multiyear stock price; noise variance; nonparametric Bayesian factor analysis; residual variance; time series clustering; Analytical models; Bayesian methods; Companies; Correlation; Data models; Load modeling; Time series analysis; Factor Analysis; Infinite Mixture Models; Nonparametric Bayesian Models; Pitman-Yor Process;
Conference_Titel :
Statistical Signal Processing Workshop (SSP), 2011 IEEE
Conference_Location :
Nice
Print_ISBN :
978-1-4577-0569-4
DOI :
10.1109/SSP.2011.5967742
