RoLNiP: Robust Learning Using Noisy Pairwise Comparisons

TL;DR

This paper introduces RoLNiP, a robust learning method for noisy pairwise comparisons that does not require noise rate knowledge in uniform noise cases and estimates noise rates in conditional noise scenarios, outperforming existing methods.

Contribution

The paper proposes a novel risk minimization framework with specific loss function conditions for robustness to noise in pairwise comparison data, including noise rate estimation techniques.

Findings

RoLNiP outperforms state-of-the-art methods in noisy pairwise comparison learning.

The approach is effective without prior noise rate knowledge in uniform noise settings.

The method accurately estimates noise rates in conditional noise scenarios.

Abstract

This paper presents a robust approach for learning from noisy pairwise comparisons. We propose sufficient conditions on the loss function under which the risk minimization framework becomes robust to noise in the pairwise similar dissimilar data. Our approach does not require the knowledge of noise rate in the uniform noise case. In the case of conditional noise, the proposed method depends on the noise rates. For such cases, we offer a provably correct approach for estimating the noise rates. Thus, we propose an end-to-end approach to learning robust classifiers in this setting. We experimentally show that the proposed approach RoLNiP outperforms the robust state-of-the-art methods for learning with noisy pairwise comparisons.