Home // DATA ANALYTICS 2015, The Fourth International Conference on Data Analytics // View article


Market Basket Analysis Using Heterogeneous Multivariate Probit Models for Groups of Product Categories

Authors:
Harald Hruschka

Keywords: Market basket analysis; Multivariate probit models, MCMC; Stochastic model search

Abstract:
Several heterogeneous multivariate probit models are used to analyze market baskets purchased by households. Each of these models is related to one group of product categories contained in seven prior partitions formed for a total of 25 product categories. The best model in terms of cross-validated log likelihood found considers all categories as one group, i.e., it does not split the 25 categories into two or more groups. In the next step of this project, we will compare this result to multivariate probit models which are related to a partition which is not fixed beforehand, but determined by stochastic model search.

Pages: 44 to 47

Copyright: Copyright (c) IARIA, 2015

Publication date: July 19, 2015

Published in: conference

ISSN: 2308-4464

ISBN: 978-1-61208-423-7

Location: Nice, France

Dates: from July 19, 2015 to July 24, 2015