Heart disease is the one of the major disease and many human beings suffered without any symptoms. In healthcare, especially in finding of heart disease in particular time plays a crucial role in cardiology area. In this paper, we proposed an effective and perfect system to predict heart disease system based on machine learning systems. This system is organized by using various classification algorithms such as LR,KNN,SVM,DT,NB and RF. The proposed algorithmic technique also solve the problem of feature selection and increase the accuracy of classification. In addition with that, the proposed algorithmic technique could use non-invasive clinical data for the heart Disease diagnosis and assessing its severity. The implementation of novel hybrid method helps to improve the accuracy of the EDA diagnosis. The proposed novel hybrid method result shows high accuracy of data is compared to previously proposed techniques. In addition to that, the proposed system is easily be adapted with  the existing technology.in this  proposed system technology more the 300 instance are collected and the results are compared with existing technology. The results prove that it has more accuracy and it can be easily implemented to identify CAD disease in healthcare field.