基于支持向量回归和核岭回归对血糖值预测的对比分析

为了对比支持向量回归(SVR)和核岭回归(KRR)预测血糖值的效果,本文选择人工智能辅助糖尿病遗传风险的相关数据进行实证分析.首先对数据进行预处理,将处理后的数据导入Python.其次,为了使SVR和KRR的对比结果具有客观性,使用了三种有代表性的核方法(线性核函数,径向基核函数和sigmod核函数).然后,在训练集上采用网格搜索自动调参分别建立SVR和KRR的最优模型,对血糖值进行预测.最后,在测试集上对比分析SVR和KRR预测的均方误差(MSE)和拟合时间等指标.结果表明:均方误差(MSE)都小于0.006,且KRR的MSE比SVR的小0.0002,KRR的预测精度比SVR更高;而SVR的预测时间比KRR的少0.803秒,SVR的预测效率比KRR好.

Contrastive Analysis of Prediction of Blood Glucose Value Based on Support Vector Regression and Kernel Ridge Regression

In order to compare the effects of support vector regression(SVR)and kernel ridge regression(KRR)in predicting blood sugar,this paper chooses the data of AI-assisted genetic risk of diabetes mellitus for empirical analysis.Firstly,the data is preprocessed,and the processed data is imported into Python.Secondly,in order to objectively compare the results of SVR and KRR,three representative kernel methods(linear kernel function,radial image kernel function and sigmod kernel function)are used.Then,the optimal models of SVR and KRR are established by grid search automatic parameter adjustment in the training set,and the blood sugar values in the test set are predicted.Finally,the MSE and fitting time of prediction errors of SVR and KRR are compared and analyzed.The results show that mean square error is less than 0.006,MSE of KRR is 0.0002 smaller than that of SVR,and the prediction accuracy of KRR is higher than that of SVR,while the prediction time of SVR is0.803 seconds less than that of KRR,and the prediction efficiency of SVR is better than that of KRR.