Trust the Unreliability: Inward Backward Dynamic Unreliability Driven Coreset Selection for Medical Image Classification

2026-03-18

AI Summary

This work addresses the challenge posed by large intra-class variability and high inter-class similarity in medical image datasets, which often undermines the effectiveness of conventional coreset selection methods. To this end, the authors propose a Dynamic Unreliability-driven Coreset Selection (DUCS) strategy that innovatively treats sample “unreliability”—characterized by fluctuations in prediction confidence and frequency of forgetting during training—as a key indicator of informativeness. By integrating introspective self-awareness with retrospective memory tracking, DUCS dynamically quantifies both sample uncertainty and the model’s memory retention, prioritizing high-value samples near decision boundaries. Extensive experiments demonstrate that DUCS significantly outperforms current state-of-the-art methods across multiple public medical imaging benchmarks, maintaining superior classification performance even under high compression ratios.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Stochastic Optimization
• Computer Vision: Medical and Biological Imaging

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Efficiently managing and utilizing large-scale medical imaging datasets with limited resources presents significant challenges. While coreset selection helps reduce computational costs, its effectiveness in medical data remains limited due to inherent complexity, such as large intra-class variation and high inter-class similarity. To address this, we revisit the training process and observe that neural networks consistently produce stable confidence predictions and better remember samples near class centers in training. However, concentrating on these samples may complicate the modeling of decision boundaries. Hence, we argue that the more unreliable samples are, in fact, the more informative in helping build the decision boundary. Based on this, we propose the Dynamic Unreliability-Driven Coreset Selection(DUCS) strategy. Specifically, we introduce an inward-backward unreliability assessment perspective: 1) Inward Self-Awareness: The model introspects its behavior by analyzing the evolution of confidence during training, thereby quantifying uncertainty of each sample. 2) Backward Memory Tracking: The model reflects on its training tracking by tracking the frequency of forgetting samples, thus evaluating its retention ability for each sample. Next, we select unreliable samples that exhibit substantial confidence fluctuations and are repeatedly forgotten during training. This selection process ensures that the chosen samples are near the decision boundary, thereby aiding the model in refining the boundary. Extensive experiments on public medical datasets demonstrate our superior performance compared to state-of-the-art(SOTA) methods, particularly at high compression rates.

Problem

• coreset selection
• medical image classification
• intra-class variation
• inter-class similarity
• decision boundary

Innovation

• coreset selection
• dynamic unreliability
• medical image classification
• decision boundary
• forgetting events