|
Activity Number:
|
99
|
|
Type:
|
Invited
|
|
Date/Time:
|
Monday, August 1, 2016 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Business and Economic Statistics Section
|
|
Abstract #318318
|
|
|
Title:
|
Fast Approximate Bayesian Analysis of Multivariate Count Time Series in a Marketing Application
|
|
Author(s):
|
Nalini Ravishanker* and Volodyymyr Serhiyenko and Rajkumar Venkatesan
|
|
Companies:
|
University of Connecticut and University of Connecticut and University of Virginia
|
|
Keywords:
|
Bayesian estimation ;
INLA ;
Level correlated models ;
marketing actions ;
pharmaceutical firm ;
time trend
|
|
Abstract:
|
In many application areas, there is increasing interest in modeling multivariate time series of counts on several subjects as a function of subject-specific and time-dependent covariates. We propose a level correlated model (LCM) to account for the association among the components of the response vector, as well as possible overdispersion. The flexible LCM framework allows us to combine different marginal count distributions and to build a hierarchical model for the vector time series of counts. We employ the Integrated Nested Laplace Approximation (INLA) for fast approximate Bayesian modeling using the R package INLA (r-inla.org). We illustrate by modeling the monthly prescription counts by physicians of a focal drug from a multinational pharmaceutical firm along with monthly counts of other competing drugs with sizable market share for the same therapeutic category. We discuss how clustering the physicians by the estimated time trend can effectively guide the firm's marketing actions.
|
Authors who are presenting talks have a * after their name.
