Ethics and Social Impacts of AI

Dataset Documentation for Responsible AI: Analysis of Suitability and Usage for Health Datasets

This study evaluates five dataset documentation approaches for health datasets and finds none are widely used or fully suitable, recommending a new standard with guidelines and tools.

Towards machine learning fairness in classifying multicategory causes of deaths in colorectal or lung cancer patients

The paper evaluates ML model fairness in classifying survival outcomes for colorectal and lung cancer patients across sociodemographic groups using TCGA data, finding biases and proposing methods to reduce disparities.

Systematic auditing is essential to debiasing machine learning in biology

This paper introduces a systematic auditing framework to identify biases in machine learning applications in biology, showing that biases can hinder model generalizability and lead to poor performance when training data lacks sufficient signal.