Self-training improves model accuracy by refining pseudo-labels.
problem Improving semi-supervised learning with self-training.
method Theoretical insights into self-training algorithm with a focus on linear classifiers.
result Self-training iterations can improve model accuracy even if stuck in sub-optimal fixed points.
Since the creation of Generative Adversarial Networks (GANs), much work has been done to improve their training stability, their generated image quality, their range of application but nearly none of them explored their self-training potential. Self-training has been used before the advent of deep learning in order to …
Doubly robust self-training improves semi-supervised learning by balancing labeled and pseudo-labeled data.
problem Improving semi-supervised learning performance with limited labeled data.
method Introduces doubly robust self-training, a method that combines labeled and pseudo-labeled data to balance between labeled-only and pseudo-labeled-only training.
result Demonstrates superior performance of doubly robust self-training on ImageNet and nuScenes datasets.
Self-training is one of the earliest and simplest semi-supervised methods. The key idea is to augment the original labeled dataset with unlabeled data paired with the model's prediction (i.e. the pseudo-parallel data). While self-training has been extensively studied on classification problems, in complex sequence gene…
Unified analysis of self-training for deep networks on unlabeled data.
problem Theoretical understanding of self-training for deep networks on unlabeled data.
method Unified theoretical analysis using expansion assumption and input-consistency regularization.
result Proves high accuracy of minimizers of population objectives based on self-training and input-consistency regularization.
Self-training outperforms pre-training on COCO object detection and segmentation datasets.
problem The effectiveness of pre-training in improving object detection and segmentation models is limited.
method Investigated self-training as an alternative method to utilize additional data.
result Self-training consistently improves model performance across various dataset sizes and data augmentation levels.
Proposes a novel graph self-training method with EM regularization for semi-supervised node classification.
problem Handles noisy graph structures and feature spaces in semi-supervised node classification.
method Introduces an Expectation-Maximization (EM) regularization scheme for uncertainty-aware pseudo-label generation and model retraining.
result Significantly outperforms strong baselines by up to 2.5% in accuracy.
GUST framework improves self-training by estimating node uncertainty and generating pseudo-labels.
problem Over-confidence in pseudo-labels during self-training.
method Graph-based uncertainty-aware self-training with stochastic node labeling.
result GUST achieves state-of-the-art performance, especially in sparse labeled data settings.
Combining self-training and contrastive learning improves performance under distribution shift.
problem Improving performance under distribution shift using unlabeled data.
method Combining self-training and contrastive learning techniques.
result Combined method achieves 3-8% higher accuracy than either approach independently.
The paper improves semi-supervised learning using f-divergences and α-Rényi divergences.
problem Improving semi-supervised learning with noisy pseudo-labels.
method Inspired by f-divergences and α-Rényi divergences, the paper develops new empirical risk functions and regularization techniques. result The new methods show better performance than traditional self-training methods, especially in noisy pseudo-label scenarios.
Self-training with noisy student-teacher boosts keyword spotting accuracy.
problem Robust keyword spotting in challenging conditions.
method Aggressive data augmentation and self-training with noisy student-teacher approach.
result Significant accuracy improvement in difficult conditions, up to 60%.
Self-training improves weak classifiers in mixture models.
problem Improving weak classifiers in mixture models.
method Iterative self-training algorithm using pseudolabels and unlabeled data.
result Self-training converts weak learners to strong learners in mixture models.
Few-shot classification (FSC) is challenging due to the scarcity of labeled training data (e.g. only one labeled data point per class). Meta-learning has shown to achieve promising results by learning to initialize a classification model for FSC. In this paper we propose a novel semi-supervised meta-learning method cal…
Proposes CST algorithm to handle partial feedback in counterfactual classification.
problem Limited feedback in settings like pricing and marketing.
method Self-training algorithm with pseudolabeling and input consistency loss.
result Improves model performance on synthetic and real datasets.
Machine learning applications in medical imaging are frequently limited by the lack of quality labeled data. In this paper, we explore the self training method, a form of semi-supervised learning, to address the labeling burden. By integrating reinforcement learning, we were able to expand the application of self train…
A new method uses optimal transport for semi-supervised classification.
problem Semi-supervised learning with limited labeled data.
method Optimal transport formulation with Sinkhorn iteration for label assignment.
result Improved performance on CIFAR-10, CIFAR-100, and SVHN datasets compared to FixMatch.
RDIS fills missing values in time series data explicitly.
problem Missing values in time series data.
method Random Drop Imputation with Self-training.
result RDIS achieves competitive results on real-world datasets.
Self-training in linear models shows a U-shaped test-risk curve due to signal forgetting and denoising.
problem Understanding the dynamics of iterative self-training in high-dimensional linear regression.
method Derivation of deterministic-equivalent recursions for prediction risk and effective noise, analysis of signal forgetting and denoising effects.
result An optimal early-stopping time is determined, and a U-shaped test-risk curve is observed.
A new confidence measure improves self-training in biased data.
problem Improving self-training in biased data.
method Proposes a new confidence measure, T-similarity, based on ensemble diversity of linear classifiers.
result Empirically shows the benefit of T-similarity for pseudo-labeling policies on various datasets.
Improved graph embedding through refined linear transformation and community recovery.
problem Identifying meaningful latent communities in graph data.
method Refined graph encoder embedding via linear transformation, self-training, and latent community recovery.
result Improved vertex embedding and better decision boundaries for vertex classification.
HMS-BERT detects cyberbullying in multiple languages and labels.
problem Multilingual and multi-label cyberbullying detection challenges.
method Hybrid multi-task self-training framework using BERT.
result Strong performance on multi-label and main classification tasks.
The paper improves self-training in semi-supervised learning by selecting more robust pseudo-labeled data.
problem Improving the reliability of pseudo-labeled data selection in self-training for semi-supervised learning.
method Proposes a multi-objective utility function to select pseudo-labeled data that maximizes reliability, considering model selection, accumulation of errors, and covariate shift uncertainties.
result Robustness towards model choice can lead to substantial accuracy gains in self-training.
Self-training algorithm improves classifier performance with labeled and unlabeled data.
problem Improving classifier performance with limited labeled data.
method Iterative learning of halfspaces, exploration and pruning phases.
result Misclassification error is bounded and never degrades compared to initial labeled set.
Machine learning systems must adapt to data distributions that evolve over time, in applications ranging from sensor networks and self-driving car perception modules to brain-machine interfaces. We consider gradual domain adaptation, where the goal is to adapt an initial classifier trained on a source domain given only…
Unsupervised framework learns latent codes for controllable generation.
problem Challenging to achieve controllable generation with GANs.
method Self-training iterative feedback from discriminator to generator.
result Better disentanglement and semantic meaningful latent codes.
Supervised Deep Learning has been highly successful in recent years, achieving state-of-the-art results in most tasks. However, with the ongoing uptake of such methods in industrial applications, the requirement for large amounts of annotated data is often a challenge. In most real world problems, manual annotation is …
Self-training improves generalization by fitting to reliable pseudo-labels or gradually improving the classification plane.
problem Understanding how self-training improves generalization in high-dimensional Gaussian mixtures.
method Analyzing iterative self-training on binary Gaussian mixtures in the asymptotic limit.
result ST improves generalization by fitting to reliable pseudo-labels or gradually improving the classification plane.
Reciprocal learning unifies various machine learning algorithms.
problem Understanding the convergence of machine learning algorithms.
method Introducing reciprocal learning as a generalization of various algorithms using decision theory.
result Reciprocal learning algorithms converge at linear rates to an approximately optimal model under certain conditions.
BOSS learns from one labeled sample per class to match fully supervised performance.
problem Achieving fully supervised performance with minimal labeled data.
method Combines class prototype refining, class balancing, and self-training.
result BOSS achieves comparable test accuracies to fully supervised learning.
Self-training avoids spurious features in domain adaptation.
problem Domain shift with large differences between source and target domains.
method Entropy minimization on unlabeled target data, initialized with a source classifier.
result Entropy minimization avoids using spurious features in large domain shifts.
In-N-Out improves model robustness to out-of-distribution data.
problem Learning robust models with few in-distribution labeled examples.
method Pre-training with auxiliary information and self-training with pseudolabels.
result In-N-Out outperforms auxiliary inputs or outputs alone on both in-distribution and OOD error.
A new approach for test-time adaptation detects and reacts to distribution shifts.
problem Improving test-time accuracy under distribution shifts.
method Online self-training with a detection tool based on entropy values and betting martingales.
result The classifier's entropy values match those of the source domain, building invariance to distribution shifts.
Improved analysis of gradual domain adaptation with better generalization bounds.
problem Improving generalization in target domain through intermediate unlabeled domains.
method Analyzed gradual self-training under more general assumptions, proving a new generalization bound.
result Proved a significantly improved generalization bound of ε0 + O(TΔ + T/√n) + ˜O(1/√nT).
Self-PU combines self-training with PU learning for improved binary classification.
problem Binary classification from unlabeled data and few labeled positives.
method Integrates self-paced training, self-calibrated loss, and self-distillation.
result Significantly improved performance on Alzheimer's Disease classification.
ASTRA uses unlabeled data and weak rules to train deep models effectively.
problem Learning with weak supervision rules is challenging due to their heuristic and noisy nature.
method ASTRA framework that considers contextualized representations and pseudo-labels for unlabeled data, and a rule attention network to aggregate labels.
result Significant improvements over state-of-the-art baselines on text classification benchmarks.
We demonstrate, theoretically and empirically, that adversarial robustness can significantly benefit from semisupervised learning. Theoretically, we revisit the simple Gaussian model of Schmidt et al. that shows a sample complexity gap between standard and robust classification. We prove that unlabeled data bridges thi…
Partial label learning deals with the problem where each training instance is assigned a set of candidate labels, only one of which is correct. This paper provides the first attempt to leverage the idea of self-training for dealing with partially labeled examples. Specifically, we propose a unified formulation with pro…
A new method uses normalizing flows for gradual domain adaptation.
problem Difficulty in domain adaptation when source and target domains have a large gap.
method Proposes using normalizing flows to learn a transformation from target to Gaussian mixture distribution.
result Improves classification performance and mitigates the problem of gradual self-training failure.
A cost-effective framework for gradual domain adaptation using multifidelity.
problem Degrading prediction performance due to large domain distance.
method Combines multifidelity and active domain adaptation.
result Improves prediction performance with reduced sample cost.
In this paper we explore the role of sample mean in building a neural network for classification. This role is surprisingly extensive and includes: direct computation of weights without training, performance monitoring for samples without known classification, and self-training for unlabeled data. Experimental computat…
Graph neural networks (GNNs) are designed for semi-supervised node classification on graphs where only a subset of nodes have class labels. However, under extreme cases when very few labels are available (e.g., 1 labeled node per class), GNNs suffer from severe performance degradation. Specifically, we observe that exi…
SLEID detects illicit accounts in DeFi transactions using semi-supervised learning.
problem Detecting illicit accounts in DeFi transactions with scarce labeled data.
method SLEID uses Isolation Forest for initial detection and self-training for pseudo-labels.
result SLEID outperforms baselines with significant improvements in precision and accuracy.
Graph clustering is a fundamental task which discovers communities or groups in networks. Recent studies have mostly focused on developing deep learning approaches to learn a compact graph embedding, upon which classic clustering methods like k-means or spectral clustering algorithms are applied. These two-step framewo…
FROST speeds up and stabilizes one-shot semi-supervised learning.
problem Slow training and sensitivity to labeled data choices in semi-supervised learning.
method Combines semi-supervised learning with a one-stage, single network self-training approach.
result FROST trains up to 10x faster and is more robust to labeled data choices.
In this paper we revisit the idea of pseudo-labeling in the context of semi-supervised learning where a learning algorithm has access to a small set of labeled samples and a large set of unlabeled samples. Pseudo-labeling works by applying pseudo-labels to samples in the unlabeled set by using a model trained on the co…
We present Noisy Student Training, a semi-supervised learning approach that works well even when labeled data is abundant. Noisy Student Training achieves 88.4% top-1 accuracy on ImageNet, which is 2.0% better than the state-of-the-art model that requires 3.5B weakly labeled Instagram images. On robustness test sets, i…
OSAMD adapts online to changing distributions with limited labels.
problem Models struggle with continual distribution shifts and expensive labeling in changing environments.
method Online Active Continual Adaptation with OSAMD, an online teacher-student structure and margin-based criterion.
result OSAMD achieves favorable dynamic regret bounds under changing environments with limited labels.
GraphFL tackles semi-supervised node classification on graphs using federated learning.
problem Real-world graph-based problems often require collecting the entire graph and labeling a reasonable number of labels, which is impractical and costly.
method GraphFL is a federated learning framework that addresses non-IID data, new label domains, and unlabeled data issues in graph-based semi-supervised node classification.
result GraphFL significantly outperforms compared FL baselines and self-training methods.