Explore the critical considerations of fairness in machine learning development through this insightful ACM talk by Tulsee Doshi, Product Lead for Google's Machine Learning Fairness Effort. Delve into lessons learned from Google's products and research, and discover approaches for evaluating and mitigating common fairness concerns in AI. Learn about the importance of explainability in addressing fairness issues and gain knowledge of available tools and techniques. Examine topics such as AI principles, ML examples and concerns, gender shades, counterfactual fairness, equality of opportunity, and improvements in mitigations. Understand the significance of transparency, industry-wide conversations, and Google's responsibility in promoting fairness in machine learning.
Fairness in Machine Learning with Tulsee Doshi
Association for Computing Machinery (ACM) via YouTube
Overview
Syllabus
Introduction
ACM
Housekeeping
Overview
What is fairness
AI principles
ML example
ML concerns
Gender shades
Counterfactual Fairness
Equality of Opportunity
Improvements Mitigations
Recap
Transparency
Tools
Google Responsibility
Industrywide conversation
Questions and answers
Replicating the model
Bias in algorithms
ML fairness in sensor data
symmetric vs asymmetric data sets
subgroup analysis and fairness
closing remarks
Taught by
Association for Computing Machinery (ACM)
Related Courses
-
AI- From Algorithms to Ethics - ACM WomENcourage 2020
-
Machine Learning Fairness - Lessons Learned
-
Fairness in Machine Learning - From Theory to Practice
-
Counterfactual Fairness
-
Building Fairer AI Systems with Fairlearn - AI Show
-
Implementing Intersectionality in Algorithmic Fairness - Keynote Panel I
