Every Sample a Task: Pushing the Limits of Heterogeneous Models with Personalized Regression

Anonymous

Keywords: Regression, Personalization, Multitask Learning

TL;DR: We present a method to estimate collections of regression models in which each model is personalized to a single sample.

Abstract:

When data arise from multiple latent subpopulations, machine learning frameworks typically estimate parameter values independently for each sub-population. In this paper, we propose
to overcome these limits by considering samples as tasks in a multitask learning framework.