Title:	Predicting occupancy rate for hospital divisions
DNr:	NAISS 2024/22-366
Project Type:	NAISS Small Compute
Principal Investigator:	Oskar Holmström <oskar.holmstrom@liu.se>
Affiliation:	Linköpings universitet
Duration:	2024-03-07 – 2024-07-01
Classification:	20206
Keywords:

Abstract

This research project utilizes machine learning and deep learning techniques to predict hospital occupancy rates in Norway, aiming to optimize resource allocation and patient care. The primary purpose of this study is to facilitate more effective resource allocation, improve patient care, and optimize hospital operations by accurately forecasting occupancy levels. By analyzing historical hospital admission data, patient demographics, seasonal trends, and other relevant variables, the project seeks to identify patterns and correlations that can inform predictive models. To achieve this, the study employs a variety of ML and DL techniques, including regression analysis, time series forecasting, and neural networks, to evaluate their efficacy in predicting occupancy rates. Rigorous hyperparameter tuning and feature selection require significant computational resources to perform the study in a timely manner.