Our #83 Webinar Meetup will be hosted by André Cruz and dedicated to the topic of “Fairness-Aware Hyperparameter Optimization”. This Meetup will be held exclusively online.
The schedule for the webinar is the following:
• 18:30 – 18:45: Opening the meeting
• 18:45 – 19:30: Fairness-Aware Hyperparameter Optimization by André Cruz
• 19:30 – 19:45: Q&A
• 19:45: Closing
See you there!
This webinar is sponsored by NOS (https://www.nos.pt/). Thank you!
Fairness-Aware Hyperparameter Optimization
Considerable research effort has been guided towards algorithmic fairness but there is still no major breakthrough. In practice, an exhaustive search over all possible techniques and hyperparameters is needed to find optimal fairness-accuracy trade-offs. Hence, coupled with the lack of tools for ML practitioners, real-world adoption of bias reduction methods is still scarce.
To address this, we present Fairband, a bandit-based fairness-aware hyperparameter optimization (HO) algorithm. We enable seamless and efficient integration of fairness objectives into real-world ML pipelines. When compared with popular HO methods, Fairband consistently finds configurations that are substantially fairer at a small decrease in predictive accuracy.
MSc Informatics and Computing Engineering from FEUP (2020).
IEEE Outstanding MSc Thesis Award, in the area of Computational Intelligence.
Currently working as a Data Scientist at Feedzai, researching fair model development in ML.
Previously worked as a Teaching Assistant at FEUP, and as a Research Assistant at TUM.