Prediction of cardiovascular diseases using random forest and naive Bayes algorithm

Abstract

Cardiovascular diseases (CVDs) are usually diagnosed by specialized doctors called cardiologists. Diagnosing cardiac disorders is a very tough task and requires correct and efficient execution. If not done properly, one may incur undesired outputs that can be lethal. Furthermore, it is illogical for an atypical man to experience exorbitant tests like electrocardiogram every time he feels unwell. Thus, very clearly, CVDs continue to be the main cause of mortality and morbidity, and in consequence, its prompt diagnosis is of paramount importance. Hence, an automatic diagnostic system is quite preferable. For this purpose, we make use of the process of data mining, also known as the proficiency discovering from data. Medical data combined with machine learning algorithm makes a prediction model that can forecast the uncertainty level of these diseases for a person based on the collected inputs. Random forest classifier and naive Bayes are two such algorithms that employ diverse probabilistic, statistical and optimization methods to identify useful patterns in large, complex and unstructured datasets. Various other algorithms like support vector machine, decision tree, K-nearest neighbor and artificial neural network are also available; hence, the primary focus is to use and amalgamate different algorithms for predicting diseases using machine learning. This chapter aims at calculating various classifier evaluation measures for the prediction of CVDs using the random forest algorithm and the naive Bayes algorithm. Furthermore, it intends to inspect and examine if the results are in accordance with the literature survey done to know which algorithm most efficiently predicts the diseases.