Applying Bayesian Inference for Signal Extraction Model of Time Series
by Jae J. Lee
This paper explores a way to apply Bayesian Inference to Signal Extraction model of time series. The basic idea is to treat hyperparameters in stochastic time series models of the components as a random vector and compute a posterior density function using Bayesian inference. Then Importance Sampling as a variance reduction technique is applied to integrate out the hyperparameters. A real time series data are used to explain how to apply this concept.
Hyperparameters, Bayesian Inference, Importance Sampling, Signal Extraction, Time Series Analysis
Jae J. Lee
READING THE ARTICLE: You can read the article in
portable document (.pdf) format (94851 bytes.)
NOTE: The content of this article is the intellectual property of the authors, who retains all rights to future publication.
This page has been accessed 2687 times since July 24, 2006.
Return to the Home Page.