Machine Learning Based Heart Disease Prediction System

Abstract

Heart disease is the main cause of the enormous deaths in the world over the past few decades and it turned out to be the deadliest not only in India but all over the world. It is therefore necessary to have a reliable, precise and functional one. To diagnose these diseases in time for appropriate treatment we used machine learning algorithms techniques. Heart is larger organs than the brain, which have a higher status in the human body to pump blood and delivers it to all organs throughout the body. Data analysis is useful for making predictions based on more information. Huge amounts of patient data are stored per month. Stored data can be useful as a source for predicting future events diseases. Machine learning techniques are used to predict heart disease, such as Artificial Neural Network, Random Forest and Support Vector Machine (SVM). The main reason of demise is because of Heart assault in world. Deaths due to heart disease have become one of the major problems with roughly one person losing their life every minute due to heart disease. Machine learning, when implemented in healthcare, is capable of early and accurate disease detection for heart disease. The datasets used have attributes of medical parameters. The datasets are processed in python using the ML Algorithm. This technique uses past old patient records to get prediction of new ones in the early stages of preventing loss of life. In this work, a heart disease prediction system is implemented using a powerful Random Forest algorithm, SVM, Naive Bayes, Decision Tree, Logistic Regression. It loads the patient data record in the form of a CSV file. After accessing the data set, the operation is performed and the effective heart attack level is create. The advantage is high performance, level of accuracy. it is very flexible and has high rates. WHO surveyed 10 million people affected by heart disease. The problem facing the healthcare industry in today's life is the timely prediction of disease after a person is affected. History records are very extensive and real-world data can be incomplete, inconsistent. In the past, it was not possible to predict the treatment of the disease for every patient in the early stages. Come up with the idea of predicting heart disease with 90% Accuracy is achieved in testing now. Practical use of data collect from previous records is time-consuming. To overcome this, we implement Random Forest algorithm to get accurate results in less time. The dataset Pre-processing, we use contains NaN. Real-time application in a variety of contexts.