Fairness (machine learning)

Fairness (machine learning)

Fairness in machine learning is the effort to correct algorithmic bias in automated decision-making processes. Decisions may be deemed unfair if they are based on sensitive variables such as gender, ethnicity, sexual orientation, or disability. The issue of algorithmic bias is well-known and extensively studied in machine learning, with outcomes potentially skewed by various factors, leading to perceived unfairness towards certain groups or individuals.

1 courses cover this concept

CS 271 / BIOMEDIN 220 Artificial Intelligence in Healthcare

Stanford University

Fall 2022-2023

Offered by Stanford University, this course focuses on AI applications in healthcare, exploring deep learning models for image, text, multimodal, and time-series data in the healthcare context. Topics also address AI integration challenges like interpretability and privacy.

No concepts data

+ 27 more concepts