Fast exact leave-one-out cross-validation of sparse least-squares support vector machines
Fast exact leave-one-out cross-validation of sparse least-squares support vector machines
复制标题
DOI：
10.1016/j.neunet.2004.07.002
复制
发表时间：
2004-12-01
期刊：
影响因子：
7.8
通讯作者：
Talbot, NLC
中科院分区：
文献类型：
--
作者：
Cawley, GC;Talbot, NLC
留一法交叉验证已被证明可以对统计模型的泛化性质给出一个几乎无偏的估计，因此为模型选择和比较提供了一个合理的标准。在本文中，我们证明了稀疏最小二乘支持向量机（LS-SVM）的精确留一交叉验证可以实现，计算复杂度仅为O（ln（2））浮点运算，而不是简单实现的O（l（2）n（2））运算，其中R是训练模式的数量，n是基向量的数量。因此，留一法交叉验证成为大规模应用中模型选择的一个实用命题。为清楚起见，论述集中在稀疏最小二乘支持向量机在非线性回归的背景下，但同样适用于模式识别设置。(C)2004爱思唯尔有限公司保留所有权利。
Leave-one-out cross-validation has been shown to give an almost unbiased estimator of the generalisation properties of statistical models, and therefore provides a sensible criterion for model selection and comparison. In this paper we show that exact leave-one-out cross-validation of sparse Least-Squares Support Vector Machines (LS-SVMs) can be implemented with a computational complexity of only O(ln(2)) floating point operations, rather than the O(l(2)n(2)) operations of a naive implementation, where R is the number of training patterns and n is the number of basis vectors. As a result, leave-one-out cross-validation becomes a practical proposition for model selection in large scale applications. For clarity the exposition concentrates on sparse least-squares support vector machines in the context of non-linear regression, but is equally applicable in a pattern recognition setting. (C) 2004 Elsevier Ltd. All rights reserved.