Learning from Label Proportion with Online Pseudo-Label Decision by Regret Minimization

TL;DR

This paper introduces an online pseudo-labeling approach with regret minimization for learning from label proportions, effectively handling large bags and demonstrating strong performance on benchmark datasets.

Contribution

The paper presents a novel LLP method based on regret minimization that works efficiently with large bag sizes, improving over previous approaches.

Findings

Effective on benchmark datasets

Handles large bag sizes efficiently

Outperforms previous LLP methods

Abstract

This paper proposes a novel and efficient method for Learning from Label Proportions (LLP), whose goal is to train a classifier only by using the class label proportions of instance sets, called bags. We propose a novel LLP method based on an online pseudo-labeling method with regret minimization. As opposed to the previous LLP methods, the proposed method effectively works even if the bag sizes are large. We demonstrate the effectiveness of the proposed method using some benchmark datasets.