Big Data in Healthcare: Predictive Modeling for Patient Outcomes
Abstract
The healthcare big data analytics movement has come with the right information to enhance overall patient outcomes, use of resources, and clinical decision-making. Through past and current patient information, predictive modelling provides medical practitioners with the capability to forecast such results as the course of the illness, the response of the patient to treatment, and the chances of the patient returning to the hospital. This study aims to perform a prediction of diabetes using a dataset that integrates medical and demographic measures. To predict the risk of diabetes, the study utilizes “machine learning models,” such as “Logistic Regression” and “Random Forest”. Also, it provides solutions to the challenges, such as an imbalance in classes. The performance of the model is studied using criteria as accuracy, recall, precision, and “area under the receiver operating curve (AUCROC)”. The results show that the “Random Forest model” is more effective than the “Logistic Regression model” in predicting diabetes, based on the accuracy and the “area under the curve (AUC)” scores. The study defines the role of big data and ML in the process of enhancing predictive potential within the healthcare market.