Scholarlink

Here, we describe the primal and dual soft margin SVM problems and give the results of numerical experiments with solving simplified dual SVM problems with soft margin. We solved the resulting QP problems with SPS or SPD Hessian and bound constraints on four classification datasets from LIBSVM dataset web page [204], namely Heart, Diabetes, Australian, and Ionosphere. We used a standard MPRGP algorithm and a variant of the Barzilai and Borwain method.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Support Vector Machines

  • 章节

  • 发表日期：

    2025年

查看全文
  • Zdeněk Dostál

摘要

Here, we describe the primal and dual soft margin SVM problems and give the results of numerical experiments with solving simplified dual SVM problems with soft margin. We solved the resulting QP problems with SPS or SPD Hessian and bound constraints on four classification datasets from LIBSVM dataset web page [204], namely Heart, Diabetes, Australian, and Ionosphere. We used a standard MPRGP algorithm and a variant of the Barzilai and Borwain method.