Impact of Data Augmentation on Confidence Calibration in Melanoma Classification

2026-10-05

AI Summary

This study addresses the unclear impact of data augmentation on model confidence calibration for imbalanced datasets in melanoma classification. By training deep neural networks on the SIIM-ISIC 2020 dataset with various data augmentation techniques, this work systematically evaluates their effects on model performance through comparisons of AUC and Expected Calibration Error (ECE). To our knowledge, this is the first investigation to comprehensively elucidate the specific mechanisms by which data augmentation, as a transformation strategy, influences model calibration in imbalanced medical imaging. Experimental results demonstrate that diverse augmentation strategies significantly enhance uncertainty calibration performance. These findings provide empirical evidence supporting the development of reliable diagnostic systems in medical imaging.

Abstract

Accurately quantifying the predictive uncertainty or improving model calibration plays an important role in medical image classification, in particular in melanoma diagnosis, where accurate uncertainty quantification can have significant implications for patient care. One of the methods for calibration improvement is data augmentation. In addition, data augmentation as a method for synthetically increasing the size of the dataset has been proven to improve the performance of models trained on imbalanced datasets. However, the impact of data augmentation, as a transformation of a part of the original data, on calibration of models trained on imbalanced datasets, in particular in melanoma classification is under-explored. We train neural networks on SIIM-ISIC 2020 melanoma classification dataset under two conditions: with and without data augmentation, and compare the differences in AUC and expected calibration error (ECE) in both scenarios. Our results shows improvements in uncertainty calibration using different augmentation methods.

Problem

• Melanoma Classification
• Data Augmentation
• Confidence Calibration
• Uncertainty Quantification
• Imbalanced Datasets

Innovation

• Data Augmentation
• Confidence Calibration
• Melanoma Classification
• Uncertainty Quantification
• Imbalanced Datasets