# Longitudinal Ensemble Integration
## Overview This repository contains the development and implementation of Longitudinal EI, a time-series and multiclass extension of the Ensemble Integration Algorithm. Here, LEI was used for early diagnosis trajectory prediction on ADNI patients via TADPOLE.
## Novel Algorithm A new algorithm has been developed to address longitudinal forecasting for multi-modal patient data.
## Implementation The repository includes a comprehensive implementation of LEI on TADPOLE. The implementation demonstrates the effectiveness and efficiency of the algorithm as well as its interpretation.
## Repository Structure - eipy/: Contains the source code for LEI and its implementation. - tadpole_data/: Contains the dataset used for the implementation as well as the data preparation pipeline code. - LEI_implement/: Contains the execution of the LEI algorithm on the data
If you use eipy in a scientific publication please cite the original study.
Full citation:
Yan Chak Li, Linhua Wang, Jeffrey N Law, T M Murali, Gaurav Pandey, Integrating multimodal data through interpretable heterogeneous ensembles, Bioinformatics Advances, Volume 2, Issue 1, 2022, vbac065, https://doi.org/10.1093/bioadv/vbac065