Is the Problem in the Data? Examples on Fairness, Diversity, and Bias

Explore the complex issue of bias in machine learning models through an in-depth analysis of PAIR's AI Explorables. Delve into the ongoing debate about whether the problem lies in the data or other factors like loss functions and network architecture. Examine three key explorables: Measuring Fairness, Hidden Bias, and Measuring Diversity. Gain insights into how these interactive tools demonstrate various aspects of bias, fairness, and diversity in AI systems. Learn to critically evaluate the sources of bias in machine learning and understand the importance of addressing these issues for more equitable AI development.

- Intro & Overview
- Recap: Bias in ML
- AI Explorables
- Measuring Fairness Explorable
- Hidden Bias Explorable
- Measuring Diversity Explorable
- Conclusion & Comments