Nonparametric Regression for Learning

: In recent years, learning theory has been increasingly influenced by the fact that many learning algorithms have at least in part a comprehensive interpretation in terms of well established statistical theories. Furthermore, with little modification, several statistical methods can be directly cast into learning algorithms. One family of such methods stems from nonparametric regression. This paper compares nonparametric learning with the more widely used parametric counterparts and investigates how these two families differ in their properties and their applicability. 1 Introduction This paper will investigate learning in a very restricted sense in that it only focuses on a low level of learning, i.e., how to establish a transformation from input data to output data under a given performance criterion. In statistics, such a mapping is called regression if the output variables are continuous, and it is called classification if the outputs are discrete. In learning theory, the acquir..