UFII Fellows Journal Club Seminar Series
“Fairness Toolkits Analysis: Juvenile Justice Case Study”
by Kiana Alikhademi
Ph.D. Candidate, Computer & Information Sciences & Engineering Department
Monday, March 28, 2022, 12:00PM – 1:00PM via Zoom
Machine learning (ML) techniques are being utilized across many domains today. Despite their strength in recognizing complex patterns in the data, they are susceptible to amplifying existing biases, which leads to unfairness. Therefore, one issue that must be addressed is fairness in many ML models’ development, use, and outcomes. Any bias in a dataset or model can have significant societal impacts in domains such as criminal justice. Fairness toolkits could play a huge role in bias mitigation and detection in this space as many researchers do not necessarily have algorithmic fairness and machine learning background. In this work, the juvenile justice dataset will be used with different fairness toolkits to evaluate them based on the rubrics developed. On the other hand, this work provides a list of flaws that we believe need to be addressed to promote better experiences for researchers and new directions for future work in the field.
Kiana Alikhademi is a PhD candidate in the Computer & Information Sciences & Engineering Department at the University of Florida, advised by Dr. Juan E. Gilbert. Kiana’s research is focused on algorithmic fairness, Bias in AI, AI for Social Good, and Computer vision.
Monday, March 28, 2022
12:00PM – 1:00PM
Registration Below. Instructions to attend talk via Zoom will be emailed to you.