Pith. sign in

REVIEW 6 cited by

Theoretical Analysis of Self-Training with Deep Networks on Unlabeled Data

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2010.03622 v5 pith:Y3JFSUIP submitted 2020-10-07 cs.LG stat.ML

classification cs.LGstat.ML
keywords self-traininganalysisdatalearningnetworkstheoreticalalgorithmsdeep
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original abstract

Self-training algorithms, which train a model to fit pseudolabels predicted by another previously-learned model, have been very successful for learning with unlabeled data using neural networks. However, the current theoretical understanding of self-training only applies to linear models. This work provides a unified theoretical analysis of self-training with deep networks for semi-supervised learning, unsupervised domain adaptation, and unsupervised learning. At the core of our analysis is a simple but realistic "expansion" assumption, which states that a low probability subset of the data must expand to a neighborhood with large probability relative to the subset. We also assume that neighborhoods of examples in different classes have minimal overlap. We prove that under these assumptions, the minimizers of population objectives based on self-training and input-consistency regularization will achieve high accuracy with respect to ground-truth labels. By using off-the-shelf generalization bounds, we immediately convert this result to sample complexity guarantees for neural nets that are polynomial in the margin and Lipschitzness. Our results help explain the empirical successes of recently proposed self-training algorithms which use input consistency regularization.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Direct Diffusion Score Preference Optimization via Stepwise Contrastive Policy-Pair Supervision

    cs.CV 2025-12 conditional novelty 6.0 of 10

    Diffusion image models can be aligned without human labels by supervising every denoising step with score targets from original versus degraded prompts.

  2. Beyond Entropy: Region Confidence Proxy for Wild Test-Time Adaptation

    cs.CV 2025-05 conditional novelty 6.0 of 10

    ReCAP models each test sample's local feature neighborhood as a Gaussian and optimizes closed-form bounds on regional entropy and instability, improving wild test-time adaptation accuracy.

  3. Self-Boost via Optimal Retraining: An Analysis via Approximate Message Passing

    cs.LG 2025-05 conditional novelty 6.0 of 10

    For binary classification with noisy labels, the paper derives the Bayes-optimal function for combining a model's current predictions with the given labels during retraining, and shows a fitted version improves linear...

  4. Relating Misfit to Gain in Weak-to-Strong Generalization Beyond the Squared Loss

    cs.LG 2025-01 conditional novelty 6.0 of 10

    For convex and approximately convex model classes, the loss gain in weak-to-strong learning is at least the KL misfit between strong and weak models, plus an error term that vanishes as k grows.

  5. Negative Metric Learning for Graphs

    cs.LG 2025-05 reject novelty 5.0 of 10

    NML-GCL learns a negative metric network that down-weights false negatives during contrastive training, improving node classification and clustering on six standard graph benchmarks.

  6. Lungmix: A Mixup-Based Strategy for Generalization in Respiratory Sound Classification

    cs.SD 2024-12 conditional novelty 5.0 of 10

    Lungmix, a mixup variant with loudness masks and semantic OR label interpolation, improves some cross-dataset respiratory sound classification scores by up to 3.55 points.

Pith tools