Cardiovascular Disease Classification Based on Machine Learning Algorithms Using GridSearchCV, Cross Validation and Stacked Ensemble Methods

Abstract

Cardiac illness is a foremost research area in medicine that has recently gotten a lot of interest around the world. In the medical profession, there is an enormous volume of data that can be eviscerated and used for numerous determinations. In the prediction of cardiovascular illness, machine learning algorithms play a critical role. Many studies have been conducted throughout the years in order to deal with the early detection of diseases. Based on the clinical data, our study article determines if the patient is likely to be analyzed with cardiovascular disease. Various strategies for data preprocessing will be employed throughout this work, and performance analysis will be performed on the distinct classification algorithms in order to predict if the patient has heart disease or not. For the UCI cardiovascular dataset, the suggested study offered many forms of machine learning and deep learning approaches that will tackle the heart disease prediction challenge. In addition, the proposed model takes into account not only various machine learning algorithms, but also hyper tweaking the parameters using GridSearchCV, Cross Validation, and Stacked Ensemble approaches. The suggested technique provides a good interpretation of the model validation through accuracy, AUC, precision, recall, KS statistics, and cumulative gain, lift curve, learn curve, calibration curve, and cross validation curve in terms of relative accurateness.