Understanding & Handling Bias While Building AI/ML Systems

AI/ML systems work based on data. Data, however, is liable to be biased because of the way we source them. This has an effect on the way an ML system works. We will be exploring how to identify biases in the data. Once identified, we will be exploring techniques about how to handles these situations along with examples from real-life scenarios.

In this session you will Learn

How machine learning solutions can be biased, how to detect that there is a bias in your system, why bias is good, and how to decide if the bias in unhealthy. What strategies you can use to avoid these biases? How does it affect the performance if ML systems.

Why you should attend this session

In the times of #BLM, a lot of companies are interested in avoiding things like racial bias. But racial bias is only one of the problems with handling bias. It is important to understand how to identify, and handle biases in the ML systems

This session is meant for

ML practitioners, or those running teams who deal with AI/ML systems

Recordings of the past webinars are available on YouTube

https://www.youtube.com/playlist?list=PLM06uCiI9yR4ZET7i9DBTjoF8SwxzIlR5