mitigatingbias.ml↗

Mitigating Bias in Machine Learning discusses how practicing model developers might build fairer predictive systems, and avoid causing harm. Part I offers context (philosophical, legal, technical) and practical solutions. Part II discusses how we quantify different notions of fairness, where possible making connections with ideologies from other disciplines (discussed in part I). Part III analyses methods for mitigating bias, looking at the impact on the various metrics (discussed in part II).