Analysis of Data Mining Techniques For Heart Disease Prediction Using Machine Learning

Abstract

A Heart Complaint or Disease (HC) is a fatal experience that stops blood flow from supplying oxygen to the heart muscle, that’s when a heart attack occurs .There are no noticeable symptoms before a heart attack occurs, which is why it is so important to detect the early symptoms correctly. Through this thesis work an attempt has been made to predict heart attack by reviewing patient history and initial symptoms based on various findings. Here the important information is collected from two popular & reputed medical college name as Ziaur Rahman Medical & TMSS Medical College for searching based on different attributes divided into 1212 rows and 14 different column. Also various classifier algorithm has been use to predict the causes of observed symptoms. Significant risk factors for heart disease were identified using the first Hunt method of the dataset to be explored. LR, SVM, DT, RF,KNN Algorithms are prognosticate IHC using the most significant 13 threat factors. The vaticination accuracy of heart disease is shown from 81.89% to 96.30% with various classifier algorithms. Where the accuracy efficiency of Random Forest is 96.30%. Keywords: Classifier,Heart Complaint, Accuracy, KNN(K-Nearest Neighbor Algorithm), LR(Logistic Regression), SVM(Support Vector Machine), DT(Decision Tree),RF(Random Forest).