Financial distress prediction for US hospitals with machine learning following KDD Methodology

Hmida, Manel (2023) Financial distress prediction for US hospitals with machine learning following KDD Methodology. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration manual) Download (812kB) | Preview

Abstract

Assessing hospitals for financial distress would help management prepare for anticipating actions and tackle areas of strategic weaknesses to avoid potential future financial distress. This Paper aims on exploring how well machine learning can predict financial distress for US hospitals and identify which of the features has helped the most in the prediction. The author deployed three different models: Deep Neural Network (DNN), Extreme Gardient Boost and SVM on US hospitals using 8 years of financial reports going from 2012 to 2019. Features used include financial ratios combining solvency, profitability, efficiency, and structure soundness besides another nonaccounting measure which is the type of control of hospitals. SVM model recorded a superior performance with an accuracy of 98.5% pursued by XGBoost with an accuracy of 98% and DNN with 82%. Random forest classifier identified net profit margin, ‘type of control’ and ROA as the most significant features in predicting financial distress within hospitals.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Cosgrave, Noel UNSPECIFIED
Uncontrolled Keywords:	Financial distress (FD); Non-financial distress (NFD); machine learning; classification; financial ratios; predictive performance; significant features
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science R Medicine > Healthcare Industry Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in FinTech
Depositing User:	Tamara Malone
Date Deposited:	10 Jan 2025 17:00
Last Modified:	10 Jan 2025 17:00
URI:	https://norma.ncirl.ie/id/eprint/7307

Actions (login required)

View Item