One of the leading causes of mortality is cardiovascular disease. Since the health system produces a significant amount of data, detecting cardiovascular problems is becoming even more crucial. Internet - of - things healthcare systems provide a tough challenge. Machine learning is essential for making correct disease predictions. There has been extensive work in this domain, yet they have not effectively grasped the real potential of machine learning strategies in predicting risk in patients because they have not utilized large quantities of information. In this study, we suggest a unique method for enhancing the accuracy of coronary heart disease diagnosis by identifying significant features using machine learning strategies. Various feature groupings and many well-known classifications techniques have been employed to develop the prediction system. The main goals of the planned study are to improve feature selection and minimize the number of traits while producing improved outcomes. In this work, a better search optimization algorithm with a conceptual methodology is applied to recognize defining factors of cardiovascular diseases. The proposed technique can also be immediately put into practice in the medical world to detect heart disease.