Class-prior estimation for learning from positive and unlabeled data

DOI10.1007/S10994-016-5604-6zbMATH Open1412.68181arXiv1611.01586OpenAlexW3104154887MaRDI QIDQ2398088FDOQ2398088

Authors: Gang Niu, Masashi Sugiyama, Marthinus Christoffel du Plessis 

Publication date: 15 August 2017

Published in: Machine Learning (Search for Journal in Brave)

Abstract: We consider the problem of estimating the class prior in an unlabeled dataset. Under the assumption that an additional labeled dataset is available, the class prior can be estimated by fitting a mixture of class-wise data distributions to the unlabeled data distribution. However, in practice, such an additional labeled dataset is often not available. In this paper, we show that, with additional samples coming only from the positive class, the class prior of the unlabeled dataset can be estimated correctly. Our key idea is to use properly penalized divergences for model fitting to cancel the error caused by the absence of negative samples. We further show that the use of the penalized $L_1$-distance gives a computationally efficient algorithm with an analytic solution. The consistency, stability, and estimation error are theoretically analyzed. Finally, we experimentally demonstrate the usefulness of the proposed method.

Full work available at URL: https://arxiv.org/abs/1611.01586

Recommendations

• scientific article; zbMATH DE number 2080441
• Principled analytic classifier for positive-unlabeled learning via weighted integral probability metric
• Learning from positive and unlabeled examples
• Learning from positive and unlabeled data: a survey
• Information-theoretic representation learning for positive-unlabeled classification

zbMATH Keywords

positive and unlabeled learningclass-prior estimation

Mathematics Subject Classification ID

Classification and discrimination; cluster analysis (statistical aspects) (62H30) Learning and adaptive systems in artificial intelligence (68T05)

Cites Work

• On Information and Sufficiency
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• A least-squares approach to direct importance estimation
• Adjusting the outputs of a classifier to new a priori probabilities: A simple procedure
• Semi-supervised novelty detection
• Dual representation of \(\phi\)-divergences and applications.
• Perturbed Optimization in Banach Spaces I: A General Theory Based on a Weak Directional Constraint Qualification
• A Neyman–Pearson Approach to Statistical Learning
• Optimization Problems with Perturbations: A Guided Tour

Cited In (18)

• Positive and unlabeled relational classification through label frequency estimation
• A two-step anomaly detection based method for PU classification in imbalanced data sets
• On the stopping criteria for \(k\)-nearest neighbor in positive unlabeled time series classification problems
• Learning from positive and unlabeled examples
• Semi-supervised AUC optimization based on positive-unlabeled learning
• A graph-based approach for positive and unlabeled learning
• Bayesian logistic model for positive and unlabeled data
• Estimating labels from label proportions
• Joint feature selection and classification for positive unlabelled multi-label data using weighted penalized empirical risk minimization
• Principled analytic classifier for positive-unlabeled learning via weighted integral probability metric
• Estimating the class prior for positive and unlabelled data via logistic regression
• A boosting framework for positive-unlabeled learning
• Triply stochastic gradient method for large-scale nonlinear similar unlabeled classification
• Information-theoretic representation learning for positive-unlabeled classification
• Positive-unlabeled classification under class-prior shift: a prior-invariant approach based on density ratio estimation
• Classification from only positive and unlabeled functional data
• Learning from positive and unlabeled data: a survey
• Title not available (Why is that?)

This page was built for publication: Class-prior estimation for learning from positive and unlabeled data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2398088)