• Home
  • Search
  • Robust deep semi-supervised learning with label propagation and differential privacy
  • Cite Icon3
  • https://doi.org/10.3389/fcomp.2023.1114186Copy DOI Icon

Robust deep semi-supervised learning with label propagation and differential privacy

Show More
  • Abstract
  • Highlights & Summary
  • PDF
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

Semi-supervised learning (SSL) methods provide a powerful tool for utilizing abundant unlabeled data to strengthen standard supervised learning. Traditional graph-based SSL methods prevail in classical SSL problems for their intuitional implementation and effective performance. However, they encounter troubles when applying to image classification followed by modern deep learning, since the diffusion algorithms face the curse of dimensionality. In this study, we propose a simple and efficient SSL method, combining a graph-based SSL paradigm with differential privacy. We aim at developing coherent latent feature space of deep neural networks so that the diffusion algorithm in the latent space can give more precise predictions for unlabeled data. Our approach achieves state-of-the-art performance on the Cifar10, Cifar100, and Mini-imagenet benchmark datasets and obtains an error rate of 18.56% on Cifar10 using only 1% of all labels. Furthermore, our approach inherits the benefits of graph-based SSL methods with a simple training process and can be easily combined with any network architecture.

Loading PDF

Similar Papers
  • Research Article
  • Citations5

Semi-supervised learning via sparse model

  • Dec 01, 2013
  • Neurocomputing
  • Yu Wang +4
  • Research Article
  • Citations316

Graph-Based Semi-Supervised Learning: A Comprehensive Review.

  • Nov 01, 2023
  • IEEE Transactions on Neural Networks and Learning Systems
  • Zixing Song +3
  • Research Article
  • Citations21

Automatic image annotation by semi-supervised manifold kernel density estimation

  • Sep 11, 2013
  • Information Sciences
  • Ping Ji +3
  • Book Chapter

Image/Video Semantic Analysis by Semi-Supervised Learning

  • May 24, 2011
  • Jinhui Tang +2
  • Research Article
  • Citations186

Graph-based semi-supervised learning: A review

  • May 08, 2020
  • Neurocomputing
  • Yanwen Chong +3
  • Research Article
  • Citations17

Video semantic analysis based on structure-sensitive anisotropic manifold ranking

  • Feb 10, 2009
  • Signal Processing
  • Jinhui Tang +3
  • Conference Article

Semi-Supervised Learning with Density-Sensitive Manifold graph

  • Oct 01, 2010
  • Zheng Wang +2
  • Conference Article
  • Citations6

Particle Competition and Cooperation to Prevent Error Propagation from Mislabeled Data in Semi-supervised Learning

  • Oct 01, 2012
  • Fabricio Breve +1
  • Research Article
  • Citations5

Instance selection method for improving graph-based semi-supervised learning

  • Feb 13, 2018
  • Frontiers of Computer Science
  • Hai Wang +2
  • Research Article
  • Citations65

Semi-Supervised Concept Learning by Concept-Cognitive Learning and Concept Space

  • Jul 21, 2020
  • IEEE Transactions on Knowledge and Data Engineering
  • Yunlong Mi +3
  • Conference Article
  • Citations12

Speaker Attribution with Voice Profiles by Graph-Based Semi-Supervised Learning

  • Oct 25, 2020
  • Jixuan Wang +5
  • Research Article
  • Citations16

ALG: Adaptive low-rank graph regularization for scalable semi-supervised and unsupervised learning

  • Aug 29, 2019
  • Neurocomputing
  • Mingbo Zhao +4
  • Conference Article

Capped l_2,1-Norm Regularized Dictionary Coding for Scalable Semi-Supervised Learning

  • Nov 01, 2019
  • Jiao Liu +1
  • Research Article
  • Citations36

Multi granularity based label propagation with active learning for semi-supervised classification

  • Dec 18, 2021
  • Expert Systems with Applications
  • Shengdan Hu +2
  • Conference Article
  • Citations8

Anisotropic Manifold Ranking for Video Annotation

  • Jul 01, 2007
  • Jinhui Tang +4
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.