• Home
  • Search
  • A Maximum Entropy Framework for Semisupervised and Active Learning With Unknown and Label-Scarce Classes.
  • Cite Icon35
  • https://doi.org/10.1109/tnnls.2016.2514401Copy DOI Icon

A Maximum Entropy Framework for Semisupervised and Active Learning With Unknown and Label-Scarce Classes.

Show More
  • Abstract
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

We investigate semisupervised learning (SL) and pool-based active learning (AL) of a classifier for domains with label-scarce (LS) and unknown categories, i.e., defined categories for which there are initially no labeled examples. This scenario manifests, e.g., when a category is rare, or expensive to label. There are several learning issues when there are unknown categories: 1) it is a priori unknown which subset of (possibly many) measured features are needed to discriminate unknown from common classes and 2) label scarcity suggests that overtraining is a concern. Our classifier exploits the inductive bias that an unknown class consists of the subset of the unlabeled pool's samples that are atypical (relative to the common classes) with respect to certain key (albeit a priori unknown) features and feature interactions. Accordingly, we treat negative log- p -values on raw features as nonnegatively weighted derived feature inputs to our class posterior, with zero weights identifying irrelevant features. Through a hierarchical class posterior, our model accommodates multiple common classes, multiple LS classes, and unknown classes. For learning, we propose a novel semisupervised objective customized for the LS/unknown category scenarios. While several works minimize class decision uncertainty on unlabeled samples, we instead preserve this uncertainty [maximum entropy (maxEnt)] to avoid overtraining. Our experiments on a variety of UCI Machine learning (ML) domains show: 1) the use of p -value features coupled with weight constraints leads to sparse solutions and gives significant improvement over the use of raw features and 2) for LS SL and AL, unlabeled samples are helpful, and should be used to preserve decision uncertainty (maxEnt), rather than to minimize it, especially during the early stages of AL. Our AL system, leveraging a novel sample-selection scheme, discovers unknown classes and discriminates LS classes from common ones, with sparing use of oracle labeling.

Similar Papers
  • Research Article
  • Citations17

Dirichlet Process Based Active Learning and Discovery of Unknown Classes for Hyperspectral Image Classification

  • Aug 01, 2016
  • IEEE Transactions on Geoscience and Remote Sensing
  • Hao Wu +1
  • Conference Article
  • Citations13

A new semi-supervised approach for hyperspectral image classification with different active learning strategies

  • Jun 01, 2012
  • Inmaculada Dopido +3
  • Research Article
  • Citations19

Employing unlabeled data to improve the classification performance of SVM, and its application in audio event classification

  • Feb 17, 2016
  • Knowledge-Based Systems
  • Yan Leng +7
  • Research Article
  • Citations2

Generalized Category Discovery With Unknown Sample Generation.

  • Jan 01, 2025
  • IEEE transactions on image processing : a publication of the IEEE Signal Processing Society
  • Xiao Li +2
  • Research Article

Unsupervised Feature Selection-Driven Active Learning for Semi-Supervised Automatic ECG Analysis.

  • Jan 01, 2025
  • IEEE journal of biomedical and health informatics
  • Xiao Li +8
  • Research Article
  • Citations7

A Generic Semi-Supervised and Active Learning Framework for Biomedical Text Classification.

  • Jul 11, 2022
  • Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference
  • Christopher A Flores +1
  • Conference Article
  • Citations1

Active Learning with Spatial Distribution based Semi-Supervised Extreme Learning Machine for Multiclass Classification

  • May 01, 2019
  • Yuefan Xu +2
  • Research Article
  • Citations51

Federated-Learning-Empowered Semi-Supervised Active Learning Framework for Intrusion Detection in ZSM

  • Feb 01, 2023
  • IEEE Communications Magazine
  • Faisal Naeem +2
  • Conference Article
  • Citations39

Semi-supervised Active Learning for Semi-supervised Models: Exploit Adversarial Examples with Graph-based Virtual Labels

  • Oct 01, 2021
  • Jiannan Guo +8
  • Conference Article
  • Citations5

ActiveMatch: End-To-End Semi-Supervised Active Representation Learning

  • Oct 16, 2022
  • Xinkai Yuan +2
  • Conference Article
  • Citations3

Combining active learning and semi-supervised for improving learning performance

  • Oct 26, 2011
  • Tinghuai Ma +2
  • Research Article
  • Citations13

Semi-supervised active learning algorithm for SVMs based on QBC and tri-training

  • Nov 25, 2020
  • Journal of Ambient Intelligence and Humanized Computing
  • Hailong Xu +2
  • Research Article
  • Citations26

Cross-Domain Semi-Supervised Learning Using Feature Formulation

  • Jun 27, 2011
  • IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics)
  • Xingquan Zhu
  • Research Article
  • Citations12

Collaborative learning for hyperspectral image classification

  • Nov 23, 2017
  • Neurocomputing
  • Chao Pan +3
  • Research Article

A novel adaptive safe semi-supervised learning framework for pattern extraction and classification

  • Jan 01, 2024
  • AIMS Mathematics
  • Jun Ma +2
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.