|
Activity Number:
|
270
- Bayesian Data Science and Analytics
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2020 : 1:00 PM to 2:50 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
|
Abstract #312696
|
|
|
Title:
|
Sequential Bayesian Updating of Time Series of Proportions
|
|
Author(s):
|
Refik Soyer*
|
|
Companies:
|
George Washington University
|
|
Keywords:
|
Linear Bayes;
NonGaussian Time Series;
Beta distribution;
Proportions
|
|
Abstract:
|
Time series of proportions arise in many applications in economics, finance and marketing. Bayesian analysis of such time series poses computational challenges due to the lack of analytical forms for sequential Bayesian updating and requires the use of Markov chain Monte Carlo (MCMC) methods. MCMC methods are not computationally efficient for sequential analysis and thus, are not attractive for real-time online processing of time series of proportions. We propose a dynamic general linear model setup for analysis of proportions and develop linear Bayesian inference using a class of conjugate priors. We illustrate the implementation of our approach using actual marketing data and compare our results with previous findings.
|
Authors who are presenting talks have a * after their name.
