针对传统网络流量预测精度低难题,为了获得理想的网络流量预测结果,提出一种基于高斯过程回归(GPR)的网络流量预测模型.该模型首先计算延迟时间和嵌入维数,构建高斯过程回归的学习样本;然后采用高斯过程回归对网络流训练集进行学习,并采用入侵杂草优化对高斯过程回归的参数进行优化;最后采用经典的网络流量测试集对该模型性能进行实验测试.实验结果表明,高斯过程回归模型提高了网络流量的预测精度.

%%%%%%%%%%      Gaussian Process Regression (GPR)               %%%%%%%%%

% Demo: prediction using GPR

% ---------------------------------------------------------------------%

clc

close all

clear all

addpath(genpath(pwd))

% load data

%{

x :   training inputs

y :   training targets

xt:   testing inputs

yt:   testing targets

%}

% multiple input-multiple output

load('https://blog.csdn.net/matlab_dingdang/article/details/data/data_2.mat')

% Set the mean function, covariance function and likelihood function

% Take meanConst, covRQiso and likGauss as examples

meanfunc = @meanConst;

covfunc = @covRQiso; 

likfunc = @likGauss; 

% Initialization of hyperparameters

hyp = struct('mean', 3, 'cov', [2 2 2], 'lik', -1);

% meanfunc = [];

% covfunc = @covSEiso; 

% likfunc = @likGauss; 

% 

% hyp = struct('mean', [], 'cov', [0 0], 'lik', -1);

% Optimization of hyperparameters

hyp2 = minimize(hyp, @gp, -5, @infGaussLik, meanfunc, covfunc, likfunc,x, y);

% Regression using GPR

% yfit is the predicted mean, and ys is the predicted variance

[yfit ys] = gp(hyp2, @infGaussLik, meanfunc, covfunc, likfunc,x, y, xt);

% Visualization of prediction results

% First output

plotResult(yt(:,1), yfit(:,1))

% Second output

plotResult(yt(:,2), yfit(:,2))

[1]李振刚. 基于高斯过程回归的网络流量预测模型[J]. 计算机应用, 2014.

博主简介：擅长智能优化算法、神经网络预测、信号处理、元胞自动机、图像处理、路径规划、无人机等多种领域的Matlab仿真，相关matlab代码问题可私信交流。

部分理论引用网络文献，若有侵权联系博主删除。