CORONARY HEART DISEASE DIAGNOSIS ALGORITHM BASED ON EFFICIENT CHANNEL ATTENTION MECHANISM AND FEATURE FUSION NETWORK
Aimed at the problem of low accuracy of early diagnosis of coronary heart disease due to factors such as uncertain important features of coronary heart disease and low predictive performance of diagnostic models,a network based on efficient channel attention mechanism and feature fusion is proposed.XGBoost(eXtreme Gradient Boosting)was used to determine the important features of coronary heart disease,and a feature combination algorithm for generating images from data was designed to apply this model.In order to improve the predictive performance of diagnostic models,efficient channel attention mechanism modules and feature fusion modules were used to improve the model learning ability and feature utilization.Compared with other diagnostic algorithms on the UCI Cleveland heart disease data set,the experimental results show that the proposed algorithm is superior to traditional machine learning methods,with a prediction accuracy of 100％and good stability.
国家科技期刊平台
NSTL
万方数据
维普
