In the video below, we will combine the power of SMOTE and logistic regression to develop a model for detecting fraud. SMOTE is used for resampling purposes, while logistic regression is the algorithm we are training.