Identifying the risk of cardiovascular diseases from the analysis of physiological attributes

Abstract

In the last few years, cardiovascular diseases have been increasing at an alarming rate and in most cases, this disease has not been detected at an early stage. In our study, we have analyzed some common physiological attributes to identify a pattern among the people having a cardiovascular disease which, in further, has been used to distinguish whether a person has a risk of developing cardiovascular disease or not. To enhance the performance of the algorithm models, we have generated a secondary dataset based on the output of the classification model, pushing the accuracy of the model to 97.03%. We have also evaluated the correlation of the attributes to the chance of having cardiovascular disease and found some general observation. Producing a secondary dataset, the analysis leading to the observable patterns among the attributes and, defining general observation for cardiovascular disease using machine learning models make this study unique.