Coefficient Regularized Algorithms for Learning and Classification

摘要：

We study the learning rate for the least square regression with data dependent hypothesis and coefficient regularization algorithms based on general kernel. We give some estimates for the learning raters of both regression and classifi- cation when the hypothesis spaces are sample dependent. Under a very mild condition on the kernels we provide learning error by using a ..-functional whose rates are estimated when the target functions are in the range of the Hilbert Schmidt integral operator.

关键词： Regularized learning scheme learning rates sample dependent spaces.

作者: Gao Wenhua Sheng Baohuai Zhang Jinhua Ye Peixin

作者单位: School of Applied Mathematics, Beijing Normal university Zhuhai, Zhuhai 519087, China Department of Mathematics, Shaoxing College of Arts and Sciences, Shaoxing, Zhejiang 312000, China School of Mathematics and LPMC, Nankai University,Tianjin 300071, China

会议类型: 国际会议

会议名称: 2012 International Conference on Intelligent System Design and Engineering Applications(2012年智能系统设计与工程应用国际会议 ISDEA 2012)

会议地点: 三亚

会议语种:英文

页码: 209-211

在线出版日期: 2012-01-06（万方平台首次上网日期，不代表论文的发表时间）