Comparative Study of Linear Kernel and Gaussian Kernel in Gaussian Process Regression
Xing Yanyuan · 2023
Gaussian process regression is a powerful Bayesian theory that can effectively help predict data.In this paper, the influence of different kernel functions on the prediction results of house price data set in Gaussian process regression is studied.Linear kernel function and Gaussian kernel function are used to predict Gaussian process regression.Through comparison experiment, it is found that the choice of kernel function has great influence on the prediction result.By analyzing the convergence rate of the training set and the accuracy of the verification set, it is concluded that better results can be obtained by using linear kernel function for Gaussian process regression for this data set Index Terms-Gaussian Process Regression;Kernel function, Likehood.