Early heart disease prediction using hybrid quantum classification

The rate of heart morbidity and heart mortality increases significantly which affect the global public health and world economy. Early prediction of heart disease is crucial for reducing heart morbidity and mortality. This paper proposes two quantum machine learning methods i.e. hybrid quantum neural network and hybrid random forest quantum neural network for early detection of heart disease. The methods are applied on the Cleveland and Statlog datasets. The results show that hybrid quantum neural network and hybrid random forest quantum neural network are suitable for high dimensional and low dimensional problems respectively. The hybrid quantum neural network is sensitive to outlier data while hybrid random forest is robust on outlier data. A comparison between different machine learning methods shows that the proposed quantum methods are more appropriate for early heart disease prediction where 96.43% and 97.78% area under curve are obtained for Cleveland and Statlog dataset respectively.