This repository contains the deprecated implementation of the lexicographic split function in Weka source codes for Decision Trees and Random Trees. These implementations were used for our paper yet recently they have been extended in Python as follow.
Scikit-longitudinal (Sklong) is an open-source Python library & Scikit-Learn API compliant, tailored to longitudinal machine learning classification tasks. It is ideal for researchers, data scientists, and analysts, as it provides specialist tools for dealing with repeated-measures data challenges. Code & Documentation
As a result, they extended our introduced lexicographical optimisation approach to the following:
| Standard Model | Longitudinal-Adapted Model |
|---|---|
| Decision Tree | Lexicographical Decision Tree (API Ref.) |
| Random Forest | Lexicographical Random Forest (API Ref.) |
| Decision Tree Regressor | Lexicographical Decision Tree Regressor (API Ref.) |
| Deep Forest | Lexicographical Deep Forest (API Ref.) |
| Gradient Boosting | Lexicographical Gradient Boosting (API Ref.) |
@article{Ribeiro2024,
title = {A lexicographic optimisation approach to promote more recent features on longitudinal decision-tree-based classifiers: applications to the English Longitudinal Study of Ageing},
volume = {57},
ISSN = {1573-7462},
url = {http://dx.doi.org/10.1007/s10462-024-10718-1},
DOI = {10.1007/s10462-024-10718-1},
number = {4},
journal = {Artificial Intelligence Review},
publisher = {Springer Science and Business Media LLC},
author = {Ribeiro, Caio and Freitas, Alex A.},
year = {2024},
month = mar
}