Reduction from Complementary-Label Learning to Probability Estimates
Complementary-Label Learning (CLL) is a weakly-supervised learning problem that aims to learn a multi-class classifier from only complementary labels, which indicate a class to which an instance does not belong. Existing approaches mainly adopt the paradigm of reduction to ordinary classification, which applies specific transformations and surrogate losses to connect CLL back to ordinary classification. Those approaches, however, face several limitations, such as the tendency to overfit or be hooked on deep models. In this paper, we sidestep those limitations with a novel perspective--reduction to probability estimates of complementary classes. We prove that accurate probability estimates of complementary labels lead to good classifiers through a simple decoding step. The proof establishes a reduction framework from CLL to probability estimates. The framework offers explanations of several key CLL approaches as its special cases and allows us to design an improved algorithm that is more robust in noisy environments. The framework also suggests a validation procedure based on the quality of probability estimates, leading to an alternative way to validate models with only complementary labels. The flexible framework opens a wide range of unexplored opportunities in using deep and non-deep models for probability estimates to solve the CLL problem. Empirical experiments further verified the framework's efficacy and robustness in various settings.
Code (0)
등록된 구현이 없습니다.
Tasks
Weakly-supervised LearningSimilar Papers 제목 키워드 기반
Learning Label Embeddings for Nearest-Neighbor Multi-class Classification with an Application to Speech Recognition
We consider the problem of using nearest neighbor methods to provide a conditional probability estimate, P(y|a), when the number of labels y is large and the labels share some underlying structure. We propose a method fo…
General ClassificationMulti-class Classificationspeech-recognitionSpeech RecognitionProbabilistic Fusion and Calibration of Neural Speaker Diarization Models
End-to-End Neural Diarization (EEND) systems produce frame-level probabilistic speaker activity estimates, yet since evaluation focuses primarily on Diarization Error Rate (DER), the reliability and calibration of these …
Speaker DiarizationDomain AdaptationWillingness to pay, surplus and Insurance policy under dual theory
In this paper, we aims to state some proprieties of willingness to pay (WTP) for partial risk reduction and links with insurance within the dual theory of decision. In the case of partial reduction, we get as Langlais (2…
Test-time Recalibration of Conformal Predictors Under Distribution Shift Based on Unlabeled Examples
Modern image classifiers are very accurate, but the predictions come without uncertainty estimates. Conformal predictors provide uncertainty estimates by computing a set of classes containing the correct class with a use…
Conformal PredictionPrediction IntervalsDiversity Enhanced Active Learning with Strictly Proper Scoring Rules
We study acquisition functions for active learning (AL) for text classification. The Expected Loss Reduction (ELR) method focuses on a Bayesian estimate of the reduction in classification error, recently updated with Mea…
Active LearningDiversitytext-classificationText Classification