A python implementation of Ensemble of Classifier Chains

This project aims to improve the implementation of the Ensemble of Classifier Chains method for multi-label classification problems. ECC is a method that extends the Classifier Chains approach by incorporating an ensemble of base classifiers.

Plotting hybrid partitions for multi-label classification

This project contains code for visualizing hybrid partitions, a method that plays a significant role in my PhD thesis. The code is designed to plot data partitions, specifically to highlight the differences between label clusters and instance clusters.

Paired analysis of multi-label classification models

A paired comparison of methods in machine learning refers to a direct comparison between two models or algorithms across multiple tasks or datasets. The goal is to determine which model performs better in a head-to-head comparison on a set of metrics, such as accuracy, precision, recall, or any othe

Multi-Label Friedman-Nemenyi Analysis

This project provides a practical implementation of the Friedman and Nemenyi statistical tests specifically tailored for multi-label classification problems. These techniques are essential for evaluating and comparing the performance of multiple algorithms.

Multi-Label Evaluation Metrics

A Python implementation of various metrics for evaluating multi-label classification models. This package offers a hierarchical organization of metrics, some implemented from scratch, while others derived from scikit-learn.

HPML-J

This code is a part of my doctoral research at PPG-CC/DC/UFSCar. HPML-J is the name of the first experiment carried out: Hybrid Partitions for Multi-Label Classification with index Jaccard.

Community Detection Methods for Multi-Label Classification

Code for the Paper Award in Bracis 2023. Best Paper Runner Up.

Analyzing multi-label classification models with Win-Tie-Loss plots

A Win-Tie-Loss chart is a visual tool for comparing the performance of different algorithms or methods across multiple tasks or datasets. This type of chart summarizes how often a method "wins," "ties," or "loses" compared to other methods based on a specific performance metric.

Analyzing multi-label classification models with Wilcoxon

The Wilcoxon Test is a comprehensive set of R scripts designed for conducting non-parametric Wilcoxon tests across multiple algorithms.