analiza aiqbench
This study demonstrates that demographic bias in skin lesion classification stems from distinct mechanisms: sex bias is driven by data imbalance, while age bias consistently favors younger cohorts regardless of dataset distribution. Although adversarial and reinforcing learning strategies effectively mitigate bias in balanced or female-majority datasets, they show limited efficacy in male-majority settings, suggesting that sex-specific training data may be necessary for equitable performance. Practitioners should prioritize targeted mitigation strategies and cross-dataset validation to address domain shifts that exacerbate these demographic disparities.
Effect of Demographic Bias on Skin Lesion Classification
arXiv:2606.03214v1 Announce Type: new Abstract: In this study, we evaluate the performance of skin lesion classification using ResNet-based convolutional models, focusing on the impact of demographic bias in training data, particularly variations in patient sex and age. We use linear programming to generate datasets with controlled demographic characteristics, allowing systematic investigation of bias effects. Three learning strategies are evaluated: a single-task model, a reinforcing multi-task model, and an adversarial learning scheme. Our sex-based analysis indicates that sex-specific training datasets optimise model performance. Notably, including male patients in the training data improved performance for the male subgroup, even in female-majority cases. Reinforcing and adversarial learning schemes narrowed or eliminated bias gaps in balanced and female-majority datasets. However, these strategies proved less effective in male-majority settings, where models continued to perform better for males than females. The two learning schemes showed marginal bias reduction compared to the baseline model in predominantly male patient populations. Age-based analysis demonstrates comparable baseline performance across the three model approaches, with performance declining across age categories. Younger groups consistently achieve the highest performance, regardless of training data distribution. Although balanced training yields optimal results for the youngest age category, performance decreases in older categories. We find that sex biases arise mainly from data imbalances, while age biases consistently favour younger groups regardless of distribution. These distinct mechanisms require targeted mitigation strategies. Additionally, cross-dataset validation on two external datasets revealed that domain shifts notably affect performance and patterns of demographic bias.
Abstract:In this study, we evaluate the performance of skin lesion classification using ResNet-based convolutional models, focusing on the impact of demographic bias in training data, particularly variations in patient sex and age. We use linear programming to generate datasets with controlled demographic characteristics, allowing systematic investigation of bias effects. Three learning strategies are evaluated: a single-task model, a reinforcing multi-task model, and an adversarial learning scheme. Our sex-based analysis indicates that sex-specific training datasets optimise model performance. Notably, including male patients in the training data improved performance for the male subgroup, even in female-majority cases. Reinforcing and adversarial learning schemes narrowed or eliminated bias gaps in balanced and female-majority datasets. However, these strategies proved less effective in male-majority settings, where models continued to perform better for males than females. The two learning schemes showed marginal bias reduction compared to the baseline model in predominantly male patient populations. Age-based analysis demonstrates comparable baseline performance across the three model approaches, with performance declining across age categories. Younger groups consistently achieve the highest performance, regardless of training data distribution. Although balanced training yields optimal results for the youngest age category, performance decreases in older categories. We find that sex biases arise mainly from data imbalances, while age biases consistently favour younger groups regardless of distribution. These distinct mechanisms require targeted mitigation strategies. Additionally, cross-dataset validation on two external datasets revealed that domain shifts notably affect performance and patterns of demographic bias.
Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Computers and Society (cs.CY); Machine Learning (cs.LG)
arXiv:2606.03214 [cs.AI]
(or arXiv:2606.03214v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2606.03214 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
https://melba-journal.org/2026:011
https://doi.org/10.59275/j.melba.2026-4156 Focus to learn more DOI(s) linking to related resources
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 356 cuvinte
Subiecte conexe
MultiFair: Multimodal Balanced Fairness-Aware Medical Classification with Dual-Level Gradient Modulation
arXiv · 29 oct. 2026
How well do routinely collected demographic and clinical variables aid point-of-care lung ultrasound TB classification
arXiv · 6 oct. 2026
Personalization Without Stereotyping: Do Speech LLMs Know When to Turn a Deaf Ear?
arXiv · 12 oct. 2026
Surrogate Losses for Fairness-Aware Classification: Fisher Consistency and Calibration
arXiv · 3 nov. 2026