Learning with Multiple Complementary Labels

Abstract

A complementary label (CL) simply indicates an incorrect class of an example, but learning with CLs results in multi-class classifiers that can predict the correct class. Unfortunately, the problem setting only allows a single CL for each example, which notably limits its potential since our labelers may easily identify multiple CLs (MCLs) to one example. In this paper, we propose a novel problem setting to allow MCLs for each example and two ways for learning with MCLs. In the first way, we design two wrappers that decompose MCLs into many single CLs, so that we could use any method for learning with CLs. However, the supervision information that MCLs hold is conceptually diluted after decomposition. Thus, in the second way, we derive an unbiased risk estimator; minimizing it processes each set of MCLs as a whole and possesses an estimation error bound. We further improve the second way into minimizing properly chosen upper bounds. Experiments show that the former way works well for learning with MCLs but the latter is even better.

Corrected typos in Lemma 2, accepted by ICML 2020

References in corpus (8)

• Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
• Are Anchor Points Really Indispensable in Label-Noise Learning?
• Masking: A New Perspective of Noisy Supervision
• On the Minimal Supervision for Training Any Binary Classifier from Only Unlabeled Data
• IMAE for Noise-Robust Learning: Mean Absolute Error Does Not Treat Examples Equally and Gradient Magnitude's Variance Matters
• Mitigating Overfitting in Supervised Classification from Two Unlabeled Datasets: A Consistent Risk Correction Approach
• Online Multiclass Classification Based on Prediction Margin for Partial Feedback
• Generative-Discriminative Complementary Learning

Cited by in corpus (2)

• Complementary to Multiple Labels: A Correlation-Aware Correction Approach
• Bridging Ordinary-Label Learning and Complementary-Label Learning