基于深度神经网络的糖尿病预测分析与研究

Analysis and Research of Diabetes Prediction Based on Deep Neural Network

在大数据的时代背景下,人工智能技术不断发展和普及,机器学习技术越来越多地被运用到医疗领域中。该文基于TensorFlow框架构建深度神经网络算法诊断糖尿病,并与KNN模型、逻辑回归模型、高斯贝叶斯模型、SVM模型、随机森林模型、AdaBoost模型和XGBoost模型七种传统机器学习模型从模型准确率、精确率、召回率、AUC值和F1-Score五项指标进行对比分析,发现深度神经网络算法所构建的预测模型准确率最高,更适合于糖尿病预测问题的分析研究。

In the context of the era of big data,artificial intelligence technology continues to develop and spread,and machine learning technology is increasingly being used in the medical field.This paper builds a deep neural network algorithm for diagnosing diabetes based on the TensorFlow framework,and works with KNN models,logistic regression models,Gaussian Bayesian models,SVM models,random forest models,AdaBoost models,and XGBoost models.Comparative analysis of five indicators:accuracy rate,recall rate,AUC value and F1-Score,found that the prediction model constructed by the deep neural network algorithm has the highest accuracy rate,which is more suitable for the analysis and research of diabetes prediction problems.