Nowadays, heart disease has become a major disease among the people irrespective of the age. We are seeing this even in children dying due to the heart disease. If we can predict this even before they die, there may be huge chances of surviving. Everybody has various qualities of beat rate (pulse rate) and circulatory strain (blood pressure). We are living in a period of data. Due to the rise in the technology, the amount of data that is generated is increasing daily. Some terabytes of data are being produced and stored. For example, the huge amount of data about the patients is produced in the hospitals such as chest pain, heart rate, blood pressure, pulse rate etc. If we can get this data and apply some machine learning techniques, we can reduce the probability of people dying. In this paper we have done survey using different classification and grouping strategies, for example, KNN, Decision tree classifier, Gaussian Naïve Bayes, Support vector machine, Linear regression, Logistic regression, Random forest classifier, Random forest regression, linear descriptive analysis. We have taken the 14 attributes that are present in the dataset as an input and applying on the dataset which is taken from the UCI repository to develop and accurate model of predicting the heart disease contains colossal (huge) therapeutic (medical) information. In the proposed research, the exhibition of the conclusion model is acquired by using utilizing classification strategies. In this paper proposed an accuracy model to predict whether a person has coronary disease or not. This is implemented by comparing the accuracies of different machine-learning strategies such as KNN, Decision tree classifier, Gaussian Naïve Bayes, SVM, Logistic regression, Random forest classifier, Linear regression, Random forest regression, linear descriptive analysis