Machine Learning Models and Medical Data Diagnosis: A Comparative Analysis

Machine learning models have increasingly transformed the medical diagnosis field providing intelligent mechanisms to analyse complex medical data and assist healthcare providers. These models use algorithms that learn iteratively from historical and real-time patient data for predicting, classifying, or identifying the explanatory variable's importance. Integration of ML models with medical diagnosis systems deals with efficient healthcare challenges like early detection of diseases, accuracy in prognosis and personalised treatment planning. ML Models analyse large-scale medical data, which includes patient’s medical records, imaging reports, genetic information and symptoms to generate latent patterns and correlations that might not be possible using traditional approaches. ML models have shown significant promise in the early detection and risk assessment of cardiovascular diseases (CVD), a leading cause of global mortality. The foremost intent of this research was to assess four ML models: Support Vector Classifier (SVC), XGBoost, Logistic Regression (LR) and Random Forest (RF) for CVD prediction and diagnosis. Among these the most impressive attainment was RF, which had an accuracy of about 85%, closely followed by the SVC model. Nevertheless, the other models also yielded comparable results on the same dataset.