Refining Heart Disease Diagnosis with Machine Learning: Techniques for Optimal Medical Outcomes

Studying disease in healthcare, diagnosis occurs. Diseases are diseases that cause pain, illness, or death. Everyone deserves optimal health, according to the WHO. Healthcare professionals should do regular health exams. Globally, heart disease kills ${3 1 \%}$ of people. In low-income nations, diagnostic facilities, skilled physicians, and other variables that impact heart disease prognosis are lacking, hence many cardiac illnesses go misdiagnosed and untreated Diabetics, hypertensives, and high cholesterol make heart disease diagnosis difficult. Disruptive cardiac rhythms may cause pulmonary fissures, jugular vein thereby affecting their and mild moderate to severe edema. Cardiovascular disease has several symptoms, requiring careful treatment. Otherwise, the heart may suffer. The US College of Cardiology reports 26 million people with heart disease and 3.6 million examined annually. Most heart disease patients die within 4-5 years, however $15-35 \%$ die annually. Physical and mental illness may affect a person’s lifestyle. The pathological procedure investigates illness origins. Modern medical assistance software employs algorithms and artificial intelligence to detect cardiovascular disease early. A difficult medical endeavor is predicting cardiac disease. The researchers employed LDA, RF, GBC, DT, SVM, KNN, and SFS to predict cardiac disease. For the purpose of verification, the system utilizes K-fold cross-validation. Studies were been out to compare these six methods. With the use of two datasets, we assess how well the model performs. In both instances, the Random Forest (RE) model outperforms all other ML models. When using Sequential Selection of Features (SFS), RE is able to get a flawless accuracy rate of ${1 0 0 \%}$. In the absence of SFS, the reliability of RE is ${9 8. 1 4 \%}$. Accordingly, the most accurate and reliable model is the RE model that incorporates SFS.