Year: 2024
Type: preprint
Abstract: Semi-supervised learning (SSL) algorithms often struggle to perform well when trained on imbalanced data. In such scenarios, the generated pseudo-labels tend to exhibit a bias toward the majority class, and models relying on these pseudo-labels can further amplify this bias. Existing imbalanced SSL algorithms explore pseudo-labeling strategies based on either pseudo-label refinement (PLR) or threshold adjustment (THA), aiming to mitigate the bias through heuristic-driven designs. However, through a careful statistical analysis, we find that existing strategies are suboptimal: most PLR algorithms are either overly empirical or rely on the unrealistic assumption that models remain well-calibrated throughout training, while most THA algorithms depend on flawed metrics for pseudo-label selection. To address these shortcomings, we first derive the theoretically optimal form of pseudo-labels under class imbalance. This foundation leads to our key contribution: SEmi-supervised learning with pseudo-label optimization based on VALidation data (SEVAL), a unified framework that learns both PLR and THA parameters from a class-balanced subset of training data. By jointly optimizing these components, SEVAL adapts to specific task requirements while ensuring per-class pseudo-label reliability. Our experiments demonstrate that SEVAL outperforms state-of-the-art SSL methods, producing more accurate and effective pseudo-labels across various imbalanced SSL scenarios while remaining compatible with diverse SSL algorithms. The code is publicly available (https://github.com/ZerojumpLine/SEVAL).
Source: arXiv (Cornell University)
Authors: Li, Zeju, Zheng, Ying-Qiu, Chen, Chen, Jbabdi, Saad
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Awards: G1080657738, Human Genetics and Disease Biology: Core Renewal for the Wellcome Trust Centre for Human Genetics, Wellcome Centre for Integrative Neuroimaging, A cross-species, cross-modal approach to computational neuroanatomy, Integrative imaging of brain structure and function in populations and individuals, G5219700674, Human Genetics and Disease Biology: Core Renewal for the Wellcome Trust Centre for Human Genetics, Wellcome Centre for Integrative Neuroimaging
OpenAlex ID: W4400480495