Improving self-training under distribution shifts via anchored confidence with theoretical guarantees

• URL: http://arxiv.org/abs/2411.00586v1
• Date: Fri, 01 Nov 2024 13:48:11 GMT
• Title: Improving self-training under distribution shifts via anchored confidence with theoretical guarantees
• Authors: Taejong Joo, Diego Klabjan,
• Abstract summary: Self-training often falls short under distribution shifts due to an increased discrepancy between prediction confidence and actual accuracy. We develop a principled method to improve self-training under distribution shifts based on temporal consistency. Our approach consistently improves self-training performances by 8% to 16% across diverse distribution shift scenarios without a computational overhead.
• Score: 13.796664304274643
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Self-training often falls short under distribution shifts due to an increased discrepancy between prediction confidence and actual accuracy. This typically necessitates computationally demanding methods such as neighborhood or ensemble-based label corrections. Drawing inspiration from insights on early learning regularization, we develop a principled method to improve self-training under distribution shifts based on temporal consistency. Specifically, we build an uncertainty-aware temporal ensemble with a simple relative thresholding. Then, this ensemble smooths noisy pseudo labels to promote selective temporal consistency. We show that our temporal ensemble is asymptotically correct and our label smoothing technique can reduce the optimality gap of self-training. Our extensive experiments validate that our approach consistently improves self-training performances by 8% to 16% across diverse distribution shift scenarios without a computational overhead. Besides, our method exhibits attractive properties, such as improved calibration performance and robustness to different hyperparameter choices.

Related papers

• Bridging the Theoretical Gap in Randomized Smoothing [43.95721606759696]
  This paper introduces a new framework that bridges the gap between theoretical certified robustness and empirical accuracy. Our approach tightens the bounds of certified robustness, offering a more accurate reflection of model robustness in practice.
  arXiv  Detail & Related papers  (2025-04-03T09:05:49Z)
• Uncertainty-Aware Graph Self-Training with Expectation-Maximization Regularization [2.743479615751918]
  We propose a novel emphuncertainty-aware graph self-training approach for semi-supervised node classification. Our method incorporates an uncertainty mechanism during pseudo-label generation and model retraining. Our framework is designed to handle noisy graph structures and feature spaces more effectively.
  arXiv  Detail & Related papers  (2025-03-26T21:52:21Z)
• Protected Test-Time Adaptation via Online Entropy Matching: A Betting Approach [14.958884168060097]
  We present a novel approach for test-time adaptation via online self-training. Our approach combines concepts in betting martingales and online learning to form a detection tool capable of reacting to distribution shifts. Experimental results demonstrate that our approach improves test-time accuracy under distribution shifts while maintaining accuracy and calibration in their absence.
  arXiv  Detail & Related papers  (2024-08-14T12:40:57Z)
• Improving Adaptive Conformal Prediction Using Self-Supervised Learning [72.2614468437919]
  We train an auxiliary model with a self-supervised pretext task on top of an existing predictive model and use the self-supervised error as an additional feature to estimate nonconformity scores. We empirically demonstrate the benefit of the additional information using both synthetic and real data on the efficiency (width), deficit, and excess of conformal prediction intervals.
  arXiv  Detail & Related papers  (2023-02-23T18:57:14Z)
• Adaptive Label Smoothing with Self-Knowledge in Natural Language Generation [16.878277421402945]
  We propose a regularization scheme that brings dynamic nature into the smoothing parameter. A model in training self-regulates the extent of smoothing on the fly during forward propagation.
  arXiv  Detail & Related papers  (2022-10-22T11:52:38Z)
• SmoothMix: Training Confidence-calibrated Smoothed Classifiers for Certified Robustness [61.212486108346695]
  We propose a training scheme, coined SmoothMix, to control the robustness of smoothed classifiers via self-mixup. The proposed procedure effectively identifies over-confident, near off-class samples as a cause of limited robustness. Our experimental results demonstrate that the proposed method can significantly improve the certified $ell$-robustness of smoothed classifiers.
  arXiv  Detail & Related papers  (2021-11-17T18:20:59Z)
• Distribution Mismatch Correction for Improved Robustness in Deep Neural Networks [86.42889611784855]
  normalization methods increase the vulnerability with respect to noise and input corruptions. We propose an unsupervised non-parametric distribution correction method that adapts the activation distribution of each layer. In our experiments, we empirically show that the proposed method effectively reduces the impact of intense image corruptions.
  arXiv  Detail & Related papers  (2021-10-05T11:36:25Z)
• Distributionally Robust Learning with Stable Adversarial Training [34.74504615726101]
  Machine learning algorithms with empirical risk minimization are vulnerable under distributional shifts. We propose a novel Stable Adversarial Learning (SAL) algorithm that leverages heterogeneous data sources to construct a more practical uncertainty set.
  arXiv  Detail & Related papers  (2021-06-30T03:05:45Z)
• Learning Prediction Intervals for Regression: Generalization and Calibration [12.576284277353606]
  We study the generation of prediction intervals in regression for uncertainty quantification. We use a general learning theory to characterize the optimality-feasibility tradeoff that encompasses Lipschitz continuity and VC-subgraph classes. We empirically demonstrate the strengths of our interval generation and calibration algorithms in terms of testing performances compared to existing benchmarks.
  arXiv  Detail & Related papers  (2021-02-26T17:55:30Z)
• Learning Calibrated Uncertainties for Domain Shift: A Distributionally Robust Learning Approach [150.8920602230832]
  We propose a framework for learning calibrated uncertainties under domain shifts. In particular, the density ratio estimation reflects the closeness of a target (test) sample to the source (training) distribution. We show that our proposed method generates calibrated uncertainties that benefit downstream tasks.
  arXiv  Detail & Related papers  (2020-10-08T02:10:54Z)
• Stable Adversarial Learning under Distributional Shifts [46.98655899839784]
  Machine learning algorithms with empirical risk minimization are vulnerable under distributional shifts. We propose Stable Adversarial Learning (SAL) algorithm that leverages heterogeneous data sources to construct a more practical uncertainty set.
  arXiv  Detail & Related papers  (2020-06-08T08:42:34Z)
• BERT Loses Patience: Fast and Robust Inference with Early Exit [91.26199404912019]
  We propose Patience-based Early Exit as a plug-and-play technique to improve the efficiency and robustness of a pretrained language model. Our approach improves inference efficiency as it allows the model to make a prediction with fewer layers.
  arXiv  Detail & Related papers  (2020-06-07T13:38:32Z)
• Robust Sampling in Deep Learning [62.997667081978825]
  Deep learning requires regularization mechanisms to reduce overfitting and improve generalization. We address this problem by a new regularization method based on distributional robust optimization. During the training, the selection of samples is done according to their accuracy in such a way that the worst performed samples are the ones that contribute the most in the optimization.
  arXiv  Detail & Related papers  (2020-06-04T09:46:52Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.