Precision Healthcare for UTIs: Leveraging Machine Learning to Reduce Readmissions

Applied Computational Intelligence and Soft Computing, Journal Year: 2025, Volume and Issue: 2025(1)

Published: Jan. 1, 2025

Hospital readmissions impose a significant financial strain on healthcare systems and can adversely affect patients. Unfortunately, traditional approaches to predicting frequently lack accuracy. This presents critical challenge, as identifying patients at high risk for readmission is essential implementing preventive measures. The study introduces novel method that employs machine learning automatically extract features from patient data, eliminating labor‐intensive manual feature engineering. primary goal develop predictive models unplanned UTI Jordan University within 3 months postdischarge. executed through retrospective analysis of electronic health records January 2020 June 2023. By leveraging techniques, the identifies high‐risk by evaluating demographic, clinical, outcome characteristics, ensuring model reliability thorough optimization, validation, performance assessment. Three were developed follows: gradient‐boosting classifier (GBC), logistic regression (LR), stochastic gradient descent (SGD). GBC, SGD, LR achieved impressive accuracy rates 99%, 95%, 89%, providing strong confidence in methodology. study’s findings reveal key factors associated with readmissions, enhancing our understanding this process offering valuable framework improving care, optimizing resource allocation, supporting evidence‐based decision‐making management.

Language: Английский

Optimisation of Data Flow Control Policies under Software Defined Network Architecture for Complex Network Environments

Applied Mathematics and Nonlinear Sciences, Journal Year: 2024, Volume and Issue: 9(1)

Published: Jan. 1, 2024

Abstract In recent years, with the rapid growth of Internet-related services, traditional software-defined network architecture has gradually failed to adapt user demands and services. This paper proposes an ant colony algorithm (ACO)-based data flow control policy optimization scheme specifically designed for networks (SDNs). It been found that ACO is prone overfitting during process policies SDN, a pheromone updating strategy introduced optimize this phenomenon. After solving phenomenon, based on will be formally formulated, simulation experiments used confirm effectiveness in paper. The results show paper’s higher priority than terms four evaluation metrics: average link throughput, utilization, round-trip delay, packet loss rate. study enables strategies under also improves utilization bring about better experience.

Language: Английский