Research of Influence Factors that Possibly Lead to Cardiovascular Disease using Machine Learning

Abstract

While previous studies have studied and demonstrated that the incidence of cardiovascular disease (CVD) is related to various factors such as hypertension, high cholesterol, smoking, diabetes, obesity, lifestyle, pregnancy and so on, there still exist many unidentified factors that are valuable to be researched on. This research tries to apply three classical machine learning algorithms to deal with the data from the Kaggle website. The dataset was compiled by Alphiree from the online Cardiovascular Diseases Risk Prediction Dataset. This dataset cited data from 2021 Behavioral Risk Factor Surveillance System (BRFSS). This study uses and processes the 5,523 records collected as data from the BRFSS in 2021 from World Health Organization (WHO). It is concluded that the BMI, Weight, Age Category, Height, Green Vegetables Consumption, Fruit Consumption and FriedPotato Consumption have relatively strong relationships with the development of CVD, while General Health, Checkup, Exercise, Skin Cancer, Other Cancer, Depression, Diabetes, Arthritis, Sex, Smoking History and Alcohol Consumption have relatively weak relationships with having a CVD. This result provides some new perspectives to study the pathogenesis and treatment of CVD and point to the way for further research afterward.