On the Power of Localized Perceptron for Label-Optimal Learning of Halfspaces with Adversarial Noise

TL;DR

This paper introduces a polynomial-time online active learning algorithm for homogeneous halfspaces with adversarial noise, achieving near-optimal label and sample complexities under certain distributional assumptions.

Contribution

It presents a novel Perceptron-like algorithm that tolerates adversarial noise and is efficient for isotropic log-concave distributions, improving upon prior methods.

Findings

Achieves near-optimal label complexity of O(d  polylog(1/psilon))

Attains sample complexity of O(d/psilon)

Handles adversarial noise with (OPT) + psilon error in the agnostic setting.

Abstract

We study {\em online} active learning of homogeneous halfspaces in $\mathbb{R}^d$ with adversarial noise where the overall probability of a noisy label is constrained to be at most $\nu$. Our main contribution is a Perceptron-like online active learning algorithm that runs in polynomial time, and under the conditions that the marginal distribution is isotropic log-concave and $\nu = \Omega(\epsilon)$, where $\epsilon \in (0, 1)$ is the target error rate, our algorithm PAC learns the underlying halfspace with near-optimal label complexity of $\tilde{O}\big(d \cdot polylog(\frac{1}{\epsilon})\big)$ and sample complexity of $\tilde{O}\big(\frac{d}{\epsilon} \big)$. Prior to this work, existing online algorithms designed for tolerating the adversarial noise are subject to either label complexity polynomial in $\frac{1}{\epsilon}$, or suboptimal noise tolerance, or restrictive marginal distributions. With the additional prior knowledge that the underlying halfspace is $s$-sparse, we obtain attribute-efficient label complexity of $\tilde{O}\big( s \cdot polylog(d, \frac{1}{\epsilon}) \big)$ and sample complexity of $\tilde{O}\big(\frac{s}{\epsilon} \cdot polylog(d) \big)$. As an immediate corollary, we show that under the agnostic model where no assumption is made on the noise rate $\nu$, our active learner achieves an error rate of $O(OPT) + \epsilon$ with the same running time and label and sample complexity, where $OPT$ is the best possible error rate achievable by any homogeneous halfspace.