arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2006.09111cs.LGstat.ML
收藏

基于LS-DC损失的统一SVM算法

Unified SVM Algorithm Based on LS-DC Loss

Xidian University(西安电子科技大学);Beijing University of Posts and Telecommunications(北京邮电大学)

Zhou Shuisheng, Zhou Wendi

摘要:

在过去的二十年中,支持向量机(Support Vector Machine,SVM)已经成为一种流行的有监督机器学习模型,并且针对不同损失(包括凸损失和非凸损失)的分类/回归问题,基于SVM模型的不同KKT条件设计了许多不同的算法。在本文中,我们提出了一种算法,可以训练不同的SVM模型在\emph{unified}计划。首先,我们介绍了\emph{LS-DC}(\textbf{l}东\textbf{s} quares类型的\textbf{d} ifference的\textbf{c} onvex)损失的定义,并表明在SVM社区中最常用的损失是LS-DC损失或可以近似为LS-DC损失。基于DCA(Difference of Convex Algorithm),我们提出了一个统一的算法,称为\emph{UniSVM},它可以求解具有任意凸或非凸LS-DC损失的SVM模型,其中只计算一个向量,特别是通过特定选择的损失。特别地,对于训练具有非凸损失的鲁棒SVM模型,UniSVM具有优于所有现有算法的优势,因为它每次迭代都具有封闭形式的解,而现有算法每次迭代都需要求解L1 SVM/L2 SVM。此外,通过核矩阵的低秩近似,UniSVM可以有效地解决大规模非线性问题。为了验证所提出的算法的有效性和可行性,我们进行了许多实验上的一些小的人工问题和一些大型的基准任务有/没有离群值的分类和回归与国家的最先进的算法进行比较。实验结果表明,UniSVM可以在更少的训练时间达到相当的性能。UniSVM最大的优点是它在Matlab中的核心代码不到10行,因此用户或研究人员很容易掌握。

英文摘要:

Over the past two decades, support vector machine (SVM) has become a popular supervised machine learning model, and plenty of distinct algorithms are designed separately based on different KKT conditions of the SVM model for classification/regression with different losses, including the convex loss or nonconvex loss. In this paper, we propose an algorithm that can train different SVM models in a \emph{unified} scheme. First, we introduce a definition of the \emph{LS-DC} (\textbf{l}east \textbf{s}quares type of \textbf{d}ifference of \textbf{c}onvex) loss and show that the most commonly used losses in the SVM community are LS-DC loss or can be approximated by LS-DC loss. Based on DCA (difference of convex algorithm), we then propose a unified algorithm, called \emph{UniSVM}, which can solve the SVM model with any convex or nonconvex LS-DC loss, in which only a vector is computed, especially by the specifically chosen loss. Particularly, for training robust SVM models with nonconvex losses, UniSVM has a dominant advantage over all existing algorithms because it has a closed-form solution per iteration, while the existing algorithms always need to solve an L1SVM/L2SVM per iteration. Furthermore, by the low-rank approximation of the kernel matrix, UniSVM can solve the large-scale nonlinear problems with efficiency. To verify the efficacy and feasibility of the proposed algorithm, we perform many experiments on some small artificial problems and some large benchmark tasks with/without outliers for classification and regression for comparison with state-of-the-art algorithms. The experimental results demonstrate that UniSVM can achieve comparable performance in less training time. The foremost advantage of UniSVM is that its core code in Matlab is less than 10 lines; hence, it can be easily grasped by users or researchers.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑