Model-specific Data Subsampling with Influence Functions

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.

References in corpus (6)

• Understanding Black-box Predictions via Influence Functions
• Randomized algorithms for matrices and data
• Consistency and robustness of kernel-based regression in convex risk minimization
• Automated Scalable Bayesian Inference via Hilbert Coresets
• Iterative Hessian sketch: Fast and accurate solution approximation for constrained least-squares
• MLBench: How Good Are Machine Learning Clouds for Binary Classification Tasks on Structured Data?