mlsanityData to 7 Oct 2026

Querying Easily Flip-flopped Samples for Deep Active Learning

arXiv 2401.09787 ↗First posted 18 Jan 2024cs.LG cs.AI stat.MLcited by 7

Not on the map. Why?

Abstract (arXiv)

Active learning is a machine learning paradigm that aims to improve the performance of a model by strategically selecting and querying unlabeled data. One effective selection strategy is to base it on the model's predictive uncertainty, which can be interpreted as a measure of how informative a sample is. The sample's distance to the decision boundary is a natural measure of predictive uncertainty, but it is often intractable to compute, especially for complex decision boundaries formed in multiclass classification tasks. To address this issue, this paper proposes the it least disagree metric (LDM), defined as the smallest probability of disagreement of the predicted label, and an estimator for LDM proven to be asymptotically consistent under mild assumptions. The estimator is computationally efficient and can be easily implemented for deep learning models using parameter perturbation. The LDM-based active learning is performed by querying unlabeled data with the smallest LDM. Experimental results show that our LDM-based active learning algorithm obtains state-of-the-art overall performance on all considered datasets and deep architectures.

Key facts. Cited by 7 papers in the map (data through 7 Oct 2026). Builds on 35 of its 88 references that are in the map. First posted on arXiv 18 Jan 2024.

Builds on

35 of 88 references.

References · 35

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,441
  3. 2015Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learningcited by 4,416
  4. 2015Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classificationcited by 3,590
  5. 2016Wide Residual Networkscited by 3,057
  6. 2015DeepFool: a simple and accurate method to fool deep neural networkscited by 1,315
  7. 2017Deep Bayesian Active Learning with Image Datacited by 812
  8. 2019Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Boundscited by 602
  9. 2014OpenML: networked science in machine learningcited by 523
  10. 2019BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learningcited by 428
  11. 2019Learning Loss for Active Learningcited by 335
  12. 2017Stochastic Gradient Descent as Approximate Bayesian Inferencecited by 298
23 more
  1. 2019Variational Adversarial Active Learningcited by 284
  2. 2018Adversarial Active Learning for Deep Networks: a Margin Based Approachcited by 155
  3. 2021Batch Active Learning at Scalecited by 127
  4. 2017Stochastic gradient descent performs variational inference, converges to limit cycles for deep networkscited by 108
  5. 2021Gone Fishing: Neural Active Learning with Fisher Embeddingscited by 88
  6. 2020Sequential Graph Convolutional Network for Active Learningcited by 69
  7. 2019Bayesian Batch Active Learning as Sparse Subset Approximationcited by 68
  8. 2022Active Learning Through a Covering Lenscited by 62
  9. 2020State-Relabeling Adversarial Active Learningcited by 60
  10. 2020Is SGD a Bayesian sampler? Well, almostcited by 50
  11. 2021Stochastic Batch Acquisition: A Simple Baseline for Deep Active Learningcited by 36
  12. 2021Low Budget Active Learning via Wasserstein Distance: An Integer Programming Approachcited by 34
  13. 2021Deep Active Learning by Leveraging Training Dynamicscited by 32
  14. 2022Making Look-Ahead Active Learning Strategies Feasible with Neural Tangent Kernelscited by 25
  15. 2018Dropout-based Active Learning for Regressioncited by 24
  16. 2022A Lagrangian Duality Approach to Active Learningcited by 24
  17. 2020Understanding the Decision Boundary of Deep Neural Networks: An Empirical Studycited by 19
  18. 2021Convergence of Uncertainty Sampling for Active Learningcited by 19
  19. 2021Active Learning in Bayesian Neural Networks with Balanced Entropy Learning Principlecited by 14
  20. 2021Nuances in Margin Conditions Determine Gains in Active Learningcited by 9
  21. 2017Influence of Resampling on Accuracy of Imbalanced Classificationcited by 8
  22. 2019Deeper Connections between Neural Networks and Gaussian Processes Speed-up Active Learningcited by 4
  23. 2017Model Selection for Anomaly Detectioncited by 1

Built on by

Cited by year

Citing papers per year: 2024: 3; 2025: 4.

7 papers.

  1. 2024Provably Neural Active Learning Succeeds via Prioritizing Perplexing Samplescited by 5
  2. 2025Enhancing Cost Efficiency in Active Learning with Candidate Set Querycited by 3
  3. 2024DCoM: Active Learning for All Learnerscited by 2
  4. 2024Parameter-Efficient Active Learning for Foundational modelscited by 1
  5. 2025Adaptation Method for Misinformation Identificationcited by 0
  6. 2025Optimizing Active Learning in Vision-Language Models via Parameter-Efficient Uncertainty Calibrationcited by 0
  7. 2025SAMOSA: Sharpness Aware Minimization for Open Set Active learningcited by 0