Estimating the Generalization in Deep Neural Networks via Sparsity

TL;DR

This paper introduces a new method to estimate the generalization gap of deep neural networks by analyzing network sparsity, using two key quantities derived from training data and a simple linear model.

Contribution

The paper proposes a novel sparsity-based approach with two key metrics and a linear model to accurately estimate the generalization gap of DNNs.

Findings

Key quantities correlate with generalization ability

Linear model accurately estimates the generalization gap

Method effective across various datasets and models

Abstract

Generalization is the key capability for deep neural networks (DNNs). However, it is challenging to give a reliable measure of the generalization ability of a DNN via only its nature. In this paper, we propose a novel method for estimating the generalization gap based on network sparsity. In our method, two key quantities are proposed first. They have close relationship with the generalization ability and can be calculated directly from the training results alone. Then a simple linear model involving two key quantities are constructed to give accurate estimation of the generalization gap. By training DNNs with a wide range of generalization gap on popular datasets, we show that our key quantities and linear model could be efficient tools for estimating the generalization gap of DNNs.