基于一维卷积神经网络的糖尿病周围神经病变预测模型研究

为了实现对糖尿病周围神经病变（DPN）的早期预防，辅助医生进行早期诊断与决策，提出了一种基于一维卷积神经网络的DPN预测模型，对原始数据进行了一系列的预处理工作以提高数据的质量，此外数据集的特征维度较高，为了进一步提高预测模型的准确性，进行了主成分分析（PCA）降维处理，通过自主学习数据的特征信息，从中挖掘其有价值的医学信息与规律，来实现DPN的预测。通过支持向量机、BP神经网络和一维卷积神经网络分别建立了DPN预测模型。实验结果表明，一维卷积神经网络模型预测效果优于其他两个模型，其准确率、召回率、F1值、AUC值分别达到了0.983、0.916、0.923、0.98。

Prediction model of diabetic peripheral neuropathy based on one-dimensional convolutional neural network

Abstract: In order to achieve the early prevention of diabetic peripheral neuropathy (DPN), and assist doctors in early diagnosis and decision-making, a prediction model of DPN based on one-dimensional convolution neural network is proposed. A series of preprocessing is carried out on the original data to improve the quality of the data. In addition, due to the high feature dimensions of the data set, principal component analysis is used to reduce the dimensions, thereby further improving the accuracy of the prediction model. Through self-learning of the feature information of the data, the useful information such as medical information and laws is mined for achieving the prediction of DPN. The DPN prediction models are established by support vector machine, BP neural network and one-dimensional convolution neural network, separately. The experimental results show that the prediction effect of one-dimensional convolutional neural network model is better than that of the other two models, and its accuracy, recall rate, F1-score and AUC values are 0.983, 0.916, 0.923 and 0.98, respectively.