On Symmetric Losses for Learning from Corrupted Labels

2019/01/27 by Charoenphakdee, Nontawat, Lee, Jongyeong, Sugiyama, Masashi · 2 citations
#FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML)

paper · doi:10.48550/arxiv.1901.09314

Abstract

This paper aims to provide a better understanding of a symmetric loss. First, we emphasize that using a symmetric loss is advantageous in the balanced error rate (BER) minimization and area under the receiver operating characteristic curve (AUC) maximization from corrupted labels. Second, we prove general theoretical properties of symmetric losses, including a classification-calibration condition, excess risk bound, conditional risk minimizer, and AUC-consistency condition. Third, since all nonnegative symmetric losses are non-convex, we propose a convex barrier hinge loss that benefits significantly from the symmetric condition, although it is not symmetric everywhere. Finally, we conduct experiments to validate the relevance of the symmetric condition.

Cited by

• Accessible, Realistic, and Fair Evaluation of Positive-Unlabeled Learning Algorithms
• Learning from Uncertain Similarity and Unlabeled Data

Related

• A Symmetric Loss Perspective of Reliable Machine Learning
• Learning with Symmetric Label Noise: The Importance of Being Unhinged
• An Exploration into why Output Regularization Mitigates Label Noise
• Making Convex Loss Functions Robust to Outliers using e-Exponentiated Transformation
• Minimizing The Misclassification Error Rate Using a Surrogate Convex Loss