Machine Learning Regularization w/Moussa Sall

Regularization is a technique used to solve overfitting problems in Machine Learning.

Overfitting typically occurs when your model performs well with your training datasets, but fails to work reliably when shown new or additional data.

Nowadays, regularization is available in most ML algorithms. It helps to choose between a complex model and simple one by using optimization.

Moussa Sall is a Data Scientist at GE Digital with a Masters degree in Statistics from the University of Cincinnati.

In this talk, Moussa will show the theory and the applications of regularization in different areas. The talk will cover Lp Norms, some loss functions, Ridge and Lasso Regression, Regularized Discriminant Analysis and a few other related techniques.

A lab will also be included using SciKit learn.

Refreshments will be provided. Don't miss it and bring a friend.