Estimation of Length of Patient Stay (LOS) based on ML Algorithms
Abstract views: 56 / PDF downloads: 35
DOI:
https://doi.org/10.59287/icriret.1401Keywords:
Length of Stay, Machine Learning, Adaboost, Random Forest, PredictionAbstract
In this study, Random Forest (RF) and Adabbost (AB) algorithms from machine learning models were used to estimate the length of stay (LOS) of patients treated in a hospital. Dependent and independent variable data from 6247 patients were used for the study. The developed ML model and the AB algorithm showed the best performance. AUC, CA, F1, Prec, Recall, and MCC values for the AB model were calculated as 0.999, 0.994, 0.994, 0.994, 0.994, and 0.991, respectively. AUC, CA, F1, Prec, Recall, and MCC values of the performance measurement values of the RF algorithm were calculated as 0.982, 0.897, 0.894, 0.900, 0.897, and 0.830, respectively. ML models were developed, a patient's LOS was calculated, and bed planning for hospital management was done efficiently with the present study.