Published on Tue Oct 20 2020

Model-specific Data Subsampling with Influence Functions

Anant Raj, Cameron Musco, Lester Mackey, Nicolo Fusi

The process of model selection is time-consuming and computationally inefficient. We develop a model-specific data subsampling strategy that improves over random sampling whenever training points have varying influence.

0
0
0
Abstract

Model selection requires repeatedly evaluating models on a given dataset and measuring their relative performances. In modern applications of machine learning, the models being considered are increasingly more expensive to evaluate and the datasets of interest are increasing in size. As a result, the process of model selection is time-consuming and computationally inefficient. In this work, we develop a model-specific data subsampling strategy that improves over random sampling whenever training points have varying influence. Specifically, we leverage influence functions to guide our selection strategy, proving theoretically, and demonstrating empirically that our approach quickly selects high-quality models.