Faster Directions for Second Order SMO
Faster Directions for Second Order SMO
复制标题
DOI：
10.1007/978-3-642-15822-3_4
复制
发表时间：
2010-09
期刊：
影响因子：
--
通讯作者：
Á. Jiménez;José R. Dorronsoro
中科院分区：
文献类型：
--
作者：
Á. Jiménez;José R. Dorronsoro
二阶SMO代表了中等规模问题的SVM训练的最新技术。在它的解决方案是通过解决一系列的子问题，优化w.r.不是一对乘法器。在本文中，我们将说明SMO如何以两阶段的方式工作，首先将有界乘数的值设置为惩罚因子C，然后继续调整无界乘数。此外，在该第二阶段期间，所选择的用于更新的对经常在算法期间重复出现。利用这一点，我们将提出一个过程，结合联合收割机以前使用的下降方向，导致在这个第二阶段的迭代少得多，也可能导致显着节省内核操作。
Second order SMO represents the state–of–the–art in SVM training for moderate size problems. In it, the solution is attained by solving a series of subproblems which are optimized w.r.t just a pair of multipliers. In this paper we will illustrate how SMO works in a two stage fashion, setting first the values of the bounded multipliers to the penalty factorCand proceeding then to adjust the non–bounded multipliers. Furthermore, during this second stage the selected pairs for update often appear repeatedly during the algorithm. Taking advantage of this, we shall propose a procedure to combine previously used descent directions that results in much fewer iterations in this second stage and that may also lead to noticeable savings in kernel operations.