ɫ����

Academics and industry experts are still not entirely on the same page when it comes to researching fairness and bias in machine learning, even though the results impact people in huge ways, such as if they can receive an organ transplant, are recognized by autonomous vehicles, or advance in the hiring process.

To help aide discussion around this hot topic, hosted a seminar and panel discussion about the work that its faculty members are doing in these areas. On Nov. 6, four faculty members affiliated with ML@GT presented their recent research that is focused on different aspects of fairness and bias.

Panelist and ) assistant professor Rachel Cummings encouraged attendees to make fairness a priority.

“This field is still so new but also so important. We need more people doing fairness research,” said Cummings.

Cummings’ presentation focused on privacy, data, and algorithmic fairness. Her colleagues Swati Gupta, an assistant professor in ISyE and Judy Hoffman, an assistant professor in the discussed the mathematics of bias and fairness and analyzing fairness in computer vision systems respectively.

“This work encourages us to get back to the basics of what we are doing and why we are doing it. Looking into how these algorithms actually affect people is huge, and we should all be thinking about the impact our work can have on all kinds of people,” said Gupta.

The session was moderated by Deven Desai, an associate professor in the .

“A goal of ML@GT’s is to develop the next generation of AI pioneers who are creating new technology that is both socially and ethically responsible, and events like these are a great way to continue to have that conversation with our students,” said Desai.

The center plans to continue hosting events like this on a variety of topics.

Watch a recording of the talk at .