On the Learnability of Rich Function Classes
β Scribed by Joel Ratsaby; Vitaly Maiorov
- Publisher
- Elsevier Science
- Year
- 1999
- Tongue
- English
- Weight
- 202 KB
- Volume
- 58
- Category
- Article
- ISSN
- 0022-0000
No coin nor oath required. For personal study only.
β¦ Synopsis
The probably approximately correct (PAC) model of learning and its extension to real-valued function classes sets a rigorous framework based upon which the complexity of learning a target from a function class using a finite sample can be computed. There is one main restriction, however, that the function class have a finite VC-dimension or scale-sensitive pseudo-dimension. In this paper we present an extension of the PAC framework with which rich function classes with possibly infinite pseudo-dimension may be learned with a finite number of examples and a finite amount of partial information. As an example we consider learning a family of infinite dimensional Sobolev classes.
π SIMILAR VOLUMES
Let G be a finite abelian group, it is a difficult and unsolved problem to find a number field F whose ideal class group is isomorphic to G. In [WAS], Corollary 3.9 and in [COR], Theorem 2, it is proved that every finite abelian group is isomorphic to a factor group of the ideal class group of some
## Abstract We consider the infima $ \hat E $(__f__) on homotopy classes of energy functionals __E__ defined on smooth maps __f__: __M^n^__ β __V^k^__ between compact connected Riemannian manifolds. If __M__ contains a subβmanifold __L__ of codimension greater than the degree of __E__ then $ \hat E