论文部分内容阅读
在处理大规模数据时,近似支持向量机及其增量式版本(ISVM)是一种比传统支持向量机更加简单而有效的分类器.但在处理高维数据时,由于ISVM通过计算矩阵的逆来更新模型参数,这使得其计算效果有待提高.针对上述问题,本文提出了基于最小二乘法的增量式方法.该增量式方法通过对矩阵运算的恒等推导,把矩阵求逆问题转变成了除法运算,得到了简单的模型参数更新公式,从而获得了和ISVM同样的预测精度,且在处理高维数据时运行效率更高.在合成数据及图像和生物数据上的试验表明该增量式方法优于ISVM方法.
Approximate SVMs and their incremental versions (ISVMs) are a simpler and more efficient classifier than traditional support vector machines when dealing with large-scale data, but when dealing with high-dimensional data, ISVMs In this paper, an incremental method based on least square method is proposed to solve the above problems.This incremental method, through the equal derivation of the matrix operation, takes the matrix inversion problem To a simple algorithm for updating the model parameters to obtain the same predictive accuracy as the ISVM and to be more efficient in processing high-dimensional data.Experiments on synthetic data and image and biological data show that the Incremental methods outperform ISVM methods.