|
Activity Number:
|
132
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 1, 2016 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics in Marketing
|
|
Abstract #318666
|
|
|
Title:
|
Psychographic Market Segmentation with Very Large Number of Behavioral Factors
|
|
Author(s):
|
Atreyee Majumder* and Tapabrata Maiti
|
|
Companies:
|
Michigan State University and Michigan State University
|
|
Keywords:
|
penalized regression ;
high-dimensional data ;
LASSO ;
elastic net ;
tuning parameter selection ;
market segmentation
|
|
Abstract:
|
This paper looks at five different world cultures to model market research data. We found that the models vary with culture clusters and depend crucially on their psychographic segmentation. High-dimensional market research data has been handled by using newly developed statistical and machine learning techniques. These advanced techniques have rarely been explored for analyzing marketing data. The paper extensively explore the Ridge, LASSO and elastic net methodology to model market survey data. Additionally, the paper runs a simulation study which illustrates the cruciality of the choice of tuning parameters in these penalized regressions. The simulation study finds an optimal method for tuning parameter selection based on information criterion. Following the simulation study a market survey questionnaire data is analyzed to predict psychographic market segmentation for different countries elaborating the cost effectiveness of such an approach and establishing the requirement of culture influenced campaigning strategies.
|
Authors who are presenting talks have a * after their name.
