基于相关向量机算法的研究与应用综述

Survey of Research and Application Based on Relevance Vector Machine

相关向量机(Relevance Vector Machine,RVM)是一种基于贝叶斯理论的稀疏概率模型,利用条件分布和极大似然的估计思想,通过核函数将低维空间的非线性问题转化为高维空间的线性问题,具有学习能力好、泛化能力强、核函数选择灵活、参数设置简单等优点。由于其出色的学习性能,已经成为当前机器学习界的研究热点。介绍了经典的相关向量机算法及其改进模型,重点评述利用相关向量机算法解决故障检测、模式识别、网络空间安全等领域的分类和预测问题的思路、方法和效果,对目前存在的问题进行总结,并对未来的研究方向进行展望。

Relevance vector machine(RVM)is a sparse probability model based on Bayesian theory.It uses the idea of conditional distribution and maximum likelihood estimation to transform the nonlinear problem in low-dimensional space into the linear problem in high-dimensional space through the kernel function method.It has the advantages of good learning ability,strong generalization ability,flexible choice of kernel function,simple parameter setting and so on.Due to its excellent learning performance,it has become a research hotspot in the current machine learning community.This article introduces the classic correlation vector machine algorithm and its improved model,re-reviews the use of correlation vector machine algorithm to solve the classification,prediction problems in the fields of fault detection,pattern recognition,cyberspace security and other ideas,methods and effects.Finally,the problem is summarized and the future research direction is prospected.