Abstract: Electronic Health Records (EHRs) are valuable for predicting patient outcomes and hospital resource demands. However, EHR data's vast scale and intricate structure present significant challenges for traditional predictive models. This paper introduces a dual prediction machine learning computational model that simultaneously forecasts patient outcomes and hospital resource utilization with improved accuracy. Using a hybrid approach that combines machine learning techniques such as artificial neural networks and support vector machines. The model effectively addresses the scale and complexity of EHR data by managing the data volume and variable relationships. Designed for concurrent execution, this model allows real-time outcome and resource predictions to run, providing high availability and reliable decision support. Trained on historical EHR data and validated for accuracy and adaptability, the model continuously learns from new patient data, accepting changes in patient demographics and hospital practices. This research has substantial implications for healthcare, enabling hospitals to make real-time predictions for resource allocation and patient care. Additionally, it can reduce costs by identifying high-risk patients early, allowing for pre-emptive interventions. This study advances EHR-based decision-making and care delivery by leveraging a concurrent model structure.
Keywords: Electronic Health Records, Machine Learning, Computational Model, Support Vector Machine
|
DOI:
10.17148/IJARCCE.2024.131006
