基于支持向量机算法优化的翻转概率算法的评估

Assessing the Flipping Probability Algorithm

现如今已有多种方法用来改善支持向量机(SVMs)算法以得到更加科学合理的分类。基于SVMs算法上的一种改进算法——翻转概率算法(Flipping Probability Algorithm)也是一个具有良好性能的分类算法,可嵌入应用于分类器中。翻转概率算法的基础是由伯明翰大学Ata Kabán博士发明的。在这篇文章中,意在介绍验证翻转概率算法并评估翻转概率算法的表现以及将其与SVMs进行对比以验证性能。

There have been many researches into improving SVMs to become better classifiers. Using the flipping probability algorithm created by Dr. Ata Kaban is a good way to show classifier performance. In this paper, flipping probability algorithm is assessed to test its performance and to make a comparison with SVMs.