市场营销系学术讲座系列第一百一十三讲

主 题：A Low-Dimension Shrinkage Approach to Choice-Based Conjoint Estimation

主讲人: Yupeng Chen (Nanyang Technological University)

时 间：2019-08-08 15:00

地 点：Room 1007, Mingde Business Building

语 言：English

Abstract：

Estimating consumers' heterogeneous preferences using choice-based conjoint (CBC) data poses a considerable modeling challenge, as the amount of information elicited from each consumer is often limited. Given the lack of individual-level information, effective information pooling across consumers becomes critical for accurate CBC estimation. In this paper, we propose a new low-dimension shrinkage approach to pooling information and modeling preference heterogeneity, in which we learn a good low-dimensional affine subspace approximation of the heterogeneity distribution and shrink the individual-level partworth estimates toward this affine subspace. We develop a tractable machine learning model to implement the low-dimension shrinkage for CBC estimation, drawing on modeling techniques in the low-rank matrix recovery literature. Using an extensive simulation experiment and two field data sets, we show the superior empirical performance of our low-dimension shrinkage model compared to alternative benchmark models.

This is a joint work with Raghu Iyengar (Wharton).

Short biography：

Yupeng Chen is an Assistant Professor of Marketing at Nanyang Technological University in Singapore. His research focuses on developing machine learning models for preference estimation and treatment effect estimation, and using field experiments to understand consumer behavior. He obtained his Ph.D. in Marketing from Wharton in 2018, Ph.D. in Operations Research from Columbia University in 2015, and B.S. in Mathematics from Peking University in 2009.