Lp-norm Sauer-Shelah lemma for margin multi-category classifiers.
J. Comput. Syst. Sci.(2017)
摘要
•An Lp-norm Sauer–Shelah lemma for margin multi-category classifiers is proved.•It is dimension-free (independent of the sample size) for every finite value of p.•The corresponding guaranteed risks depend sublinearly on the number of categories.
更多查看译文
关键词
Margin multi-category classifiers,Guaranteed risks,ϵ-Entropy,γ-Dimension,Generalized Sauer–Shelah lemmas
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络