This is a machine learning-based research project for UPMC. The objective is to write an algorithm to forecast overcrowding in a hospital emergency department. The main challenge of this project was obtaining accurate predictions from a sparse data set.
See WhitePaper.pdf
Data Items: Contains original and modified versions of the data set.
Experiment: Unorganized & unoptimized MATLAB scripts, for the purpose of experimentation.
MATLAB Implementation: Optimized and organized MATLAB code for training models.
Python Implementation: Python scripts for testing and implementing models trained in MATLAB.
Results: Contains a few graphs showing the NEDOC Score predictions for models.\