Fast classification rates without standard margin assumptions

TL;DR

This paper demonstrates that fast classification rates are achievable without standard margin assumptions by using a specific algorithm in the agnostic setting, extending previous theoretical results.

Contribution

It introduces a new learning algorithm for classification with reject options that achieves near-optimal rates under misspecification, without relying on margin assumptions.

Findings

Achieves $O(d/n \, \log(n/d))$ learning rate in agnostic setting.

Performance of the proposed algorithm is never worse than ERM.

Provides necessary and sufficient conditions for fast rates in the agnostic setting.

Abstract

We consider the classical problem of learning rates for classes with finite VC dimension. It is well known that fast learning rates up to $O\left(\frac{d}{n}\right)$ are achievable by the empirical risk minimization algorithm (ERM) if low noise or margin assumptions are satisfied. These usually require the optimal Bayes classifier to be in the class, and it has been shown that when this is not the case, the fast rates cannot be achieved even in the noise free case. In this paper, we further investigate the question of the fast rates under the misspecification, when the Bayes classifier is not in the class (also called the agnostic setting). First, we consider classification with a reject option, namely Chow's reject option model, and show that by slightly lowering the impact of hard instances, a learning rate of order $O\left(\frac{d}{n}\log \frac{n}{d}\right)$ is always achievable in the agnostic setting by a specific learning algorithm. Similar results were only known under special versions of margin assumptions. We also show that the performance of the proposed algorithm is never worse than the performance of ERM. Based on those results, we derive the necessary and sufficient conditions for classification (without a reject option) with fast rates in the agnostic setting achievable by improper learners. This simultaneously extends the work of Massart and N\'{e}d\'{e}lec (Ann. of Statistics, 2006), which studied this question in the case where the Bayesian optimal rule belongs to the class, and the work of Ben-David and Urner (COLT, 2014), which allows the misspecification but is limited to the no noise setting. Our result also provides the first general setup in statistical learning theory in which an improper learning algorithm may significantly improve the learning rate for non-convex losses.