Machine Learning Modeling: How to work with unbalanced datasets

Hugh_R_Heinsohn

For predictive modeling, we need a "balanced" dataset for the algorithm to find a signal to detect and properly predict an outcome. However, some examples of fraud and failure have so few example that our model is not able to predict the minor class. This session will focus on modeling techniques to deal with imbalanced datasets when one category in your dependent variable has too few examples.

https://fast.wistia.net/embed/iframe/mmibbct5h8?seo=true&videoFoam=false