09-07, 15:15–15:50 (Europe/Lisbon), Auditorium
In Artificial Intelligence systems, a renaissance has unfolded in recent years, captivating the imagination of many! However, a startling statistic emerges amidst the enthusiasm: merely 13% of machine learning models are deployed into production!
Indeed, the artistry of building and deploying these systems supersedes mere scientific methodology as these models dwell within the realm of complexity and are characterized by an inherent reliance on data distribution. Have you ever wondered if these models are trained upon biased data they are destined to produce haywire decisions?
In this talk, let us delve into the realm of ethicality, fairness, and unbiasedness, culminating in the seamless integration of machine learning models that embrace ethical considerations in their journey towards production using Python!
Have you ever wondered why the increasing reliance on Machine Learning systems raises concerns about fairness and bias in their data-driven decisions? What if ML models are built on skewed data or are not designed to mitigate bias - then they can perpetuate and even amplify existing inequalities and injustices!
Since there's no one-size-fits-all approach, thus, building and deploying a fair and unbiased ML system is more of an art than a science! In this talk, firstly, we will explore the challenges involved in building and deploying fair and unbiased ML systems. Secondly, we will understand the technical debts which incur while building such systems and how to investigate them. Finally, we will learn fundamental strategies and best practices for ensuring your ML models are fair, unbiased, and ethical!
Rashmi is a skilled Software Engineer at Cactus and a researcher at the University of San Francisco, working in Machine Learning. With nearly four years of experience in the IT industry, she has successfully spearheaded product ideation and led redesigns and feature development, reaching millions of users. Rashmi indulges in astronomy, capturing images of the cosmos with her telescope or unwinding with friends over board games when she is not occupied with coding. Additionally, Rashmi is a proud pet parent to her Maltese breed dog, Fluffy.