Session Title: Data Leakage- Most Ignored Problem In Machine learning
Speaker: Sandip Pani
Abstract: When you share information while training model which you shouldn’t is referred as data leakage. Beginners to experts many do this mistake of sharing information while tracing model. Sometime it happens accidentally and sometimes users not even aware about his mistake.
In this session we will discuss what is data leakage?, how it impact my model and how will you detect it? and how to prevent data leakage? I will demonstrate with various examples.
300+ sessions are now available on-demand from Data Platform Summit 2021 & 2020 at no cost. Browse all sessions.
Stay tuned, more learning coming your way.