粗糙学习过程一致收敛速度的界

Bounds of the Uniform Convergence Rate of the Learning Process Concerning Rough Samples

考虑到现实世界的不确定性,结合粗糙随机理论和粗糙统计学的知识,对粗糙随机样本情况下的统计学习理论一致收敛速度的界进行了推广,提出了粗糙学习问题的一般表示;给出了粗糙风险泛函、粗糙经验风险泛函以及粗糙经验风险最小化归纳原则等概念;最终证明了粗糙随机学习过程一致收敛速度的界。

In view of the uncertainty of the real world,rough set theory to generalize key theorem and bounds of uniform convergence rate of learning processes is taken as the basis of statistical learning theory in this paper.The setting of the rough learning problem is brought forward;rough risk function/rough empirical risk function and RERM principle are presented;bounds of uniform convergence rate of rough learning process are put forward in the end.