Design of heart failure prediction software based on CatBoost model

• 1、School of information and electronic engineering, Zhejiang Gongshang universtity,Hangzhou 310018

Abstract：At present, the problem of cardiovascular disease is the focus of public health problems in China. The prevention, early warning, early diagnosis and early intervention of cardiovascular diseases have become urgent. Therefore, prevention and early detection are of great importance to reduce cardiovascular events. In recent years, cardiovascular diseases have received increasing attention from domestic and international researchers, resulting in a number of research projects that have made important contributions to cardiovascular diseases, but cardiovascular disease risk screening remains a major challenge. Since severe heart failure can lead to patient death, it is important to make advance predictions based on patients\' clinical and laboratory data. This project has designed a prediction software that can predict the probability of the disease based on the patient\'s characteristics. The core of this project is to analyze the clinical and laboratory data of the patient by using CatBoost algorithm in machine learning and finally get a good model which can predict the probability of the patient\'s disease. the CatBoost algorithm overcomes the overfitting problem and achieves good results in both classification and regression problems. It has a better advantage in predicting cardiovascular diseases.

Keywords： Machine learning heart failure CatBoost classification problems