Various types of sampling techniques for imbalanced datasets are discussed in depth with examples and analysis. Get yourself familiar with over-/under-sampling, SMOTE, ADA-SYN, sampling with cleaning, boosting, clustering, and more. Continue reading Imbalanced Learning: sampling techniques
This article attempts to summarize the popular evaluation metrics for binary classification problems. Continue reading Binary Classification Evaluation Summary
We introduce an alternative for the ROC: the Precision-Recall curve (PR-curve), which is a more reliable measurement for the cases when Positive samples are rare. Continue reading Precision-Recall curve: an overview
The well-known ROC curve plot, the Area Under the ROC Curve (AUC), and its variants. Continue reading ROC curve and AUC: a comprehensive overview