Novel Classification and Prediction of Heart Disease using CDMA Algorithm

Abstract

Chronic illness is a long-term condition that lasts a lifetime. In most cases, immunizations and medications cannot heal them, or they do not work.  The most common chronic illnesses are heart disease. The first step in stopping the progression of these disorders is patient diagnosis and prognosis. The identification of individuals with heart disease may be made easier with the machine learning (ML) and deep learning (DL). Finding people who is at risk for these well-known illnesses is often influenced by a variety of circumstances. High precision is provided by deep learning. Machine learning, however, provides less precision. Deep learning also needs a lot of data. However, machine learning can be trained on less data. By doing so, we may determine that one technique's flaw is fixed by another. To classify and forecast heart disease, this research developed an algorithm by combining ML and DL algorithm that is Combination of Machine Learning and Deep Learning Algorithm (CMDA). The data set for the work was taken from UCI data repository. The CMDA algorithm uses the Dl4jMlpClassifier and the Support Vector Machine (SVM). The technique like stacking classifier is used to integrate above two algorithms in the CMDA. The classification method utilized Naive Bayes as a meta-classifier in the CMDA algorithm, uses a stacking classifier strategy for final prediction. After prediction finally, the CMDA method utilizes the Min-Max normalization approach to determine risk factor. According to the experimental findings, the proposed CMDA algorithm effectively classifies and forecasts heart disease and produce high results while comparing with existing methods.