A new multi-view regression approach with an application to customer wallet estimation

Motivated by the problem of customer wallet estimation, we propose a new setting for multi-view regression, where we learn a completely unobserved target (in our case, customer wallet) by modeling it as a "central link" in a directed graphical model, connecting multiple sets of observed variables. The resulting conditional independence allows us to reduce the discriminative maximum likelihood estimation problem to a convex optimization problem for parametric forms corresponding to exponential linear models. We show that under certain modeling assumptions, in particular, when we have two conditionally independent views and the noise is Gaussian, we can reduce this problem to a single least squares regression. Thus, for this specific, but widely applicable setting, the "unsupervised" multi-view problem can be solved via a simple supervised learning approach. This reduction also allows us to test the statistical independence assumptions underlying the graphical model ...

• Type: Conference
• Year: 2006
• Where: KDD
• Authors: Srujana Merugu, Saharon Rosset, Claudia Perlich
• Keywords: Customer Wallet Estimation, Data Mining, KDD 2006, Likelihood Estimation Problem, Multi-view Problem