Disseminating the Risk Factors With Enhancement in Precision Medicine Using Comparative Machine Learning Models for Healthcare Data

Abstract

The field of healthcare analytics, as an expanding discipline that integrates data analysis, machine learning, and clinical expertise, is seen to hold great promise for improving patient outcomes and the overall delivery of healthcare services. With the increasing availability of Electronic Health Records (EHRs), a wealth of healthcare data has emerged, presenting opportunities to enhance disease prediction and personalize treatments. The objective of the research is to develop and evaluate machine-learning models for predicting cancer, diabetic, diabetic retina, and heart-related outcomes using demographic and clinical data from Electronic Health Records (EHRs). Through thorough testing on diverse datasets, the study aims to assess the performance of these models in terms of accuracy, precision, and recall metrics, with the ultimate goal of advancing disease prediction and enhancing patient outcomes within the field of healthcare analytics. The proposed model demonstrates high accuracy, particularly in predicting cancer (97.080%) and diabetic (97.33%) outcomes using Support Vector Machines (SVM) and Decision Trees. Additionally, logistic regression achieves a notable accuracy of 76.521% for diabetic retina dataset, while Decision Trees exhibit 86.419% accuracy for heart-related predictions. SVM accuracy for Pima diabetic dataset stands at 79.746%. To assess the model’s performance, thorough testing was conducted on a diverse and extensive dataset, employing a combination of accuracy, precision, and recall metrics. This research represents a substantial contribution to the field of healthcare analytics, emphasizing the potential of machine learning to advance disease prediction and, ultimately, enhance patient outcomes.

Keywords

• Machine learning
• decision trees
• logistic regression
• random forest classifier
• support vector machine
• breast cancer