This study deals with effectiveness of various techniques based on machine learning in the prediction of heart. Dataset sourced from Kaggle. The study analyses the performance of multiple Machine learning models like Logistic Regression, Random Forest, XGBoost, Support Vector Machines (SVM), and Multi-Layer Perceptrons (MLP). Additionally, a Voting Classifier Combining these models is done by employing techniques to boost performance with hyperparameter tuning Applied for each individual model using Grid Search. Performance is evaluated through classification Accuracy and classification report metrics. The results reveal the Voting without hyperparameter tuning, the classifier reached the best accuracy of 95.4%. Combining the Tuned models with Bagging techniques resulted in an accuracy of 94.5%. This research highlights the important potential of ML techniques in improving heart disease prediction, which Can lead to early detection and treatment.