Havi is a data scientist at Data Science Group. She works on end-to-end Machine and Deep Learning projects in a variety of fields such as medical and finance. Havi holds an MSc in industrial engineering from Tel Aviv University. Her thesis research focused on two approaches for improving the predictive abilities of ranking ensemble. During her studies, she was a researcher at the TAU Big Data Lab. Havi is a strong believer in fairness and in making the world of technology more fair, just and overall a better place!
Havi Werbin Ofir
Givat Shmuel, Israel
Can also give an online talk/webinar
Paid only. Contact speaker for pricing!
A Strategy to Detect Unfairness in AI
Data / AI / ML, Innovation
Do AI models reflect the real world, or do they make their own decisions based on biased input? An algorithm may achieve high performance, but, is it really successful? As the world turns to more AI-based systems, it's crucial to detect and correct biases that are harmful or unfair. In this talk, Havi will present the fairness issue that has arisen in certain computer vision models. She will look at specific sensitive attributes that may cause the algorithm to discriminate and be difficult to remove from the data. She will also explain how she identifies whether the algorithm exhibits discrimination and a lack of fairness. As Victor Hugo says - “Being good is easy, what is difficult is being just”. Let's show the world how fairness works!