Research on CNN to Feature Extractionin Diseases Prediction

In the hypertension and Hyperlipidemia prediction study, due to there are some problems with the feature extraction of text-based data in physical examination data. This study proposes the feature extraction using convolutional neural network, and then the gradient lifting tree algorithm establishes the prediction model. The deviation between the prediction result and the actual result is 0.0277. The group prediction bias is 0.0394. Therefore, in the hypertension and Hyperlipidemia prediction, using the convolutional neural network for feature extraction can reduce the prediction bias and improve the prediction effect.