Cardiovascular diseases is one of the main causes of death across the globe, which results in 17.9 million fatalities per year. Heart related disorders increase expenditures on healthcare and, taken in common lower a person’s productivity. It is therefore unlikely that accurate and feasible prediction of heart-related disorders is required. Developing the most beneficial machine learning classification model-based prediction to assist physicians in predicting the severity of cardiopathy is aimed in this research based on patient clinical knowledge. Most of the time, a sophisticated combination of clinical and pathological knowledge is used to diagnose cardiopathy. Because of this characteristic, academics and clinical practitioners are very interested in accurate and cost-effective cardiopathy prediction. Globally, medical organizations gather information on a range of health-related topics. Numerous machine learning algorithms utilize this data to extract meaningful data. Utilize the dataset by combining numerous previously available datasets, including the Long Beach, Virginia, Hungarian, Cleveland, Switzerland, as well as Stalog (Heart) Data Sets databases. 1190 patient records from the US, UK, Switzerland, and Hungary make up the dataset. It has one target variable and eleven characteristics. Finally, a basic cardiopathy prediction model has been generated by comparing XGBoost, RF-random forest, DT-decision tree, logistic regression, KNN, Naive Bayes, as well as linear SVM. The models will be hyperparameter tweaked for optimal precision, and the best algorithms with the best training parameters will be combined to create a stacked ensemble model.