MATLAB Code for prediction based on Relevance Vector Machine (RVM) using ‘SB2_Release_200 toolbox’.

主要特点：

1. 用于训练和测试RVM模型的易用API

2. 多种核函数

3. 可视化模块

Main features

1. Easy-used API for training and testing RVM model

2. Multiple kinds of kernel functions

3. Visualization module

一个简单示例：

A simple application

%{

RVM application for a numerical example.

%}

clc

clear all

close all

addpath(genpath(pwd))

% generate data 产生数据

[x, y, xt, yt] = generateData;

% parameter setting 参数设置

kernel = Kernel(‘type’, ‘gauss’, ‘width’, 2);

option = struct(‘freeBasis’, ‘on’,…

‘display’, ‘on’);

% train RVM model 训练RVM模型

model = rvm_train(x, y, ‘kernel’, kernel, ‘option’, option);

% predict the test samples 预测测试样本

result = rvm_test(model,xt, yt);

% visualization 可视化

plotRelevanceVector(model, y)

plotResult(yt, result)

更多精彩文章请关注公众号：