deslib

Research implementation exploring the relationship between dataset complexity and optimal k parameter in Dynamic Classifier Selection methods.

This study presents a comparative analysis of static and dynamic ensemble learning for predicting myocardial infarction mortality. To mitigate the challenges of class imbalance, the research implements a comprehensive cost-sensitive framework that operates at the data, algorithmic, and evaluation levels.

A comparative study of Static and Dynamic Ensemble Learning for credit card fraud detection. This project addresses extreme class imbalance by integrating reweighted class and undersampling techniques to optimize detection rates in skewed financial datasets.