Key result
The random forest model with Grey Wolf Optimizer achieved the highest predictive performance for 30-day hospital readmissions in patients with diabetes, with an F1 score of 0.83 and accuracy of 0.88.
Why the study?
Previous research suggested the predictive superiority of deep learning models over traditional machine learning in forecasting diabetes readmissions, prompting a comparison of one deep learning and 10 machine learning methods.
Do machine learning models with Grey Wolf Optimizer accurately predict 30-day readmission rates in hospitalized patients with diabetes?
Population
101,766 unique inpatient encounters with a diabetes diagnosis across 130 US hospitals
Comparison
1 deep learning model vs 10 machine learning models incorporating Grey Wolf Optimizer
Design
Retrospective comparative prediction model study
Follow-up
30 days
Authors
Loading...
May support ML readmission prediction in diabetes; leaves open prospective validation before clinical use.
Do machine learning models with Grey Wolf Optimizer accurately predict 30-day readmission rates in hospitalized patients with diabetes?
Random Forest and XGBoost models combined with Grey Wolf Optimizer feature selection demonstrated high accuracy in predicting 30-day hospital readmissions for patients with diabetes.
Liu et al. (2024) studied Diabetes (n=101,766). Random forest (RF) and XGBoost models with Grey Wolf Optimizer (GWO) vs. Deep learning (LSTM) and other traditional machine learning models was evaluated on 30-day readmission rate. The random forest model with Grey Wolf Optimizer achieved the highest predictive performance for 30-day hospital readmissions in patients with diabetes, with an F1 score of 0.83 and accuracy of 0.88.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: