摘要：We study the least square regression with data dependent hypothesis and coefficient regularization algorithms based on general kernel. An explicit expression of the solution of this kernel scheme is derived. Then we provide a sample error with a decay of
关键词：Least Square Regressions; Data Depen -dent Hypothesis; Coefficient Regulari -zation; General Kernel; Learning Rate.
