Heart disease prediction and detection has long been considered as a critical issue. Early detection of heart disease is an important issue in health care services (HCS). In growing amount of health care systems, patients are offered expensive therapies and operation that is quiet expensive for developing countries. Recently, heart disease is a prominent public chronic disease, ex. it's a growing concern in the US. The main reason of these diseases are tobacco consumption, bad life style, lack of physical activity and the intake of alcohol. Therefore, there is a need for the cloud based architecture that can efficiently predict and track health information. Recently, machine learning techniques have already been established to solve clinical problem and medical diagnosis. In this study, we proposed a cloud-based 4- tier architecture that can significantly improve the prediction and monitoring of patient's health information. Hence, we used five popular supervised learning based machine learning technique for early detection of heart disease. The major purpose of this study is to examine the performance of the selected classification techniques. In addition, we use prominent evaluation criteria to observe the best performance of these machine learning techniques. Moreover, we used the ten-fold cross-validation technique to evaluate the performance of the five classifiers. The analysis results indicate that the Artificial Neural Network (ANN) achieved the highest performance of all. However, health care researchers and practitioners can obtain independent understanding from this work while selecting machine learning techniques to apply in their area.