Synthetic Augmentation in Imbalanced Learning: When It Helps, When It Hurts, and How Much to Add

• URL: http://arxiv.org/abs/2601.16120v1
• Date: Thu, 22 Jan 2026 17:15:26 GMT
• Title: Synthetic Augmentation in Imbalanced Learning: When It Helps, When It Hurts, and How Much to Add
• Abstract summary: We develop a unified statistical framework for synthetic augmentation in imbalanced learning.<n>In a local symmetry" regime, imbalance is not the dominant source of error near the balanced optimum.<n>The optimal synthetic size depends on generator accuracy and on whether the generator's residual mismatch is directionally aligned with the intrinsic-minority shift.
• Score: 4.735371232176844
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Imbalanced classification, where one class is observed far less frequently than the other, often causes standard training procedures to prioritize the majority class and perform poorly on rare but important cases. A classic and widely used remedy is to augment the minority class with synthetic examples, but two basic questions remain under-resolved: when does synthetic augmentation actually help, and how many synthetic samples should be generated? We develop a unified statistical framework for synthetic augmentation in imbalanced learning, studying models trained on imbalanced data augmented with synthetic minority samples and evaluated under the balanced population risk. Our theory shows that synthetic data is not always beneficial. In a ``local symmetry" regime, imbalance is not the dominant source of error near the balanced optimum, so adding synthetic samples cannot improve learning rates and can even degrade performance by amplifying generator mismatch. When augmentation can help (a ``local asymmetry" regime), the optimal synthetic size depends on generator accuracy and on whether the generator's residual mismatch is directionally aligned with the intrinsic majority-minority shift. This structure can make the best synthetic size deviate from naive full balancing, sometimes by a small refinement and sometimes substantially when generator bias is systematic. Practically, we recommend Validation-Tuned Synthetic Size (VTSS): select the synthetic size by minimizing balanced validation loss over a range centered near the fully balanced baseline, while allowing meaningful departures when the data indicate them. Simulations and a real sepsis prediction study support the theory and illustrate when synthetic augmentation helps, when it cannot, and how to tune its quantity effectively.

Related papers

• When Does Synthetic Data Augmentation Improve Score-Based Imbalanced Classification? [4.850565550043597]
  This paper develops a framework for characterizing when synthetic minority augmentation can improve threshold-integrated and threshold-optimized metrics.<n>We separate the effect of augmentation into two components: a change in effective class weighting and a discrepancy between the synthetic and true minority distributions.<n>We provide explicit improvement bounds quantifying the roles of approximation error, finite-sample estimation error, and synthetic distributional error.
  arXiv  Detail & Related papers  (2026-06-24T17:30:36Z)
• Towards Syn-to-Real IQA: A Novel Perspective on Reshaping Synthetic Data Distributions [74.00222571094437]
  Blind Image Quality Assessment (BIQA) has advanced significantly through deep learning, but the scarcity of large-scale labeled datasets remains a challenge.<n>We make a key observation that representations learned from synthetic datasets often exhibit a discrete and clustered pattern that hinders regression performance.<n>We introduce a novel framework SynDR-IQA, which reshapes synthetic data distribution to enhance BIQA generalization.
  arXiv  Detail & Related papers  (2026-01-01T06:11:16Z)
• Bias-Corrected Data Synthesis for Imbalanced Learning [18.33651035966011]
  Imbalanced data, where the positive samples represent only a small proportion compared to the negative samples, makes it challenging for classification problems to balance the false positive and false negative rates.<n>A common approach to addressing the challenge involves generating synthetic data for the minority group and then training classification models with both observed and synthetic data.<n>In this paper, we address the bias introduced by synthetic data and provide consistent estimators for this bias by borrowing information from the majority group.
  arXiv  Detail & Related papers  (2025-10-30T00:52:25Z)
• Concentration and excess risk bounds for imbalanced classification with synthetic oversampling [5.974778743092435]
  We develop a theoretical framework to analyze the behavior of SMOTE and related methods when classifiers are trained on synthetic data.<n>Results lead to practical guidelines for better parameter tuning of both SMOTE and the downstream learning algorithm.
  arXiv  Detail & Related papers  (2025-10-23T12:12:51Z)
• Synthetic Oversampling: Theory and A Practical Approach Using LLMs to Address Data Imbalance [16.047084318753377]
  Imbalanced classification and spurious correlation are common challenges in data science and machine learning.<n>Recent advances have proposed leveraging the flexibility and generative capabilities of large language models to generate synthetic samples.<n>This article develops novel theoretical foundations to systematically study the roles of synthetic samples in addressing imbalanced classification and spurious correlation.
  arXiv  Detail & Related papers  (2024-06-05T21:24:26Z)
• A Bias-Variance Decomposition for Ensembles over Multiple Synthetic Datasets [4.389150156866014]
  Recent studies have highlighted the benefits of generating multiple synthetic datasets for supervised learning.<n>These benefits have clear empirical support, but the theoretical understanding of them is currently very light.<n>We seek to increase the theoretical understanding by deriving bias-variance decompositions for several settings of using multiple synthetic datasets.
  arXiv  Detail & Related papers  (2024-02-06T13:20:46Z)
• Let's Synthesize Step by Step: Iterative Dataset Synthesis with Large Language Models by Extrapolating Errors from Small Models [69.76066070227452]
  *Data Synthesis* is a promising way to train a small model with very little labeled data. We propose *Synthesis Step by Step* (**S3**), a data synthesis framework that shrinks this distribution gap. Our approach improves the performance of a small model by reducing the gap between the synthetic dataset and the real data.
  arXiv  Detail & Related papers  (2023-10-20T17:14:25Z)
• Synthetic data, real errors: how (not) to publish and use synthetic data [86.65594304109567]
  We show how the generative process affects the downstream ML task. We introduce Deep Generative Ensemble (DGE) to approximate the posterior distribution over the generative process model parameters.
  arXiv  Detail & Related papers  (2023-05-16T07:30:29Z)
• ContraNeRF: Generalizable Neural Radiance Fields for Synthetic-to-real Novel View Synthesis via Contrastive Learning [102.46382882098847]
  We first investigate the effects of synthetic data in synthetic-to-real novel view synthesis. We propose to introduce geometry-aware contrastive learning to learn multi-view consistent features with geometric constraints. Our method can render images with higher quality and better fine-grained details, outperforming existing generalizable novel view synthesis methods in terms of PSNR, SSIM, and LPIPS.
  arXiv  Detail & Related papers  (2023-03-20T12:06:14Z)
• Boosting Differentiable Causal Discovery via Adaptive Sample Reweighting [62.23057729112182]
  Differentiable score-based causal discovery methods learn a directed acyclic graph from observational data. We propose a model-agnostic framework to boost causal discovery performance by dynamically learning the adaptive weights for the Reweighted Score function, ReScore.
  arXiv  Detail & Related papers  (2023-03-06T14:49:59Z)
• An Embarrassingly Simple Baseline for Imbalanced Semi-Supervised Learning [103.65758569417702]
  Semi-supervised learning (SSL) has shown great promise in leveraging unlabeled data to improve model performance. We consider a more realistic and challenging setting called imbalanced SSL, where imbalanced class distributions occur in both labeled and unlabeled data. We study a simple yet overlooked baseline -- SimiS -- which tackles data imbalance by simply supplementing labeled data with pseudo-labels.
  arXiv  Detail & Related papers  (2022-11-20T21:18:41Z)
• A Synthetic Over-sampling method with Minority and Majority classes for imbalance problems [0.0]
  We propose a new method to generate synthetic instances using Minority and Majority classes (SOMM) SOMM generates synthetic instances diversely within the minority data space. It updates the generated instances adaptively to the neighbourhood including both classes.
  arXiv  Detail & Related papers  (2020-11-09T03:39:56Z)

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.