Comparative Assessment of Machine Learning Methods for Early Prediction of Diseases Using Health Indicators

Abstract

In today's modern world, traditional systems frequently find it difficult to keep up with the need for early disease identification. To effectively address these issues, early diagnoses of diseases become imperative. This work aims to develop a comprehensive platform utilizing machine learning (ML) for predictive analysis and early diagnoses of diseases. By allowing early disease identification and encouraging a holistic approach to well-being, this study aims to maximize the potential of technology to improve public health outcomes. The primary focus is on predicting heart disease, diabetes, and breast cancer using machine learning algorithms. Specifically, the Naïve Bayes (NB) classifier achieves an 85% accuracy in predicting heart disease. Support vector machine (SVM) attains 75% accuracy for diabetes prediction. The logistic regression (LR) algorithm stands out by achieving an exceptional accuracy of 98% in predicting cancer. The results highlight a high level of reliability and platform's potential for eastly diagnosis of disease.