Title :
A mathematical foundation on Poggio´s regularization theory
Author :
Watanabe, Kôtarô ; Namatame, Akira ; Kashiwaghi, Eiichi
Author_Institution :
Dept. of Comput. Sci., Nat. Defense Acad., Yokosuka, Japan
Abstract :
This paper justifies the regularization theory for learning of an input-output mapping from a set of examples, which was proposed by Poggio. Moreover, the relationships between the dimension of input space, the restriction condition and the smoothness of the map obtained as a solution are proposed.
Keywords :
Hilbert spaces; learning (artificial intelligence); neural nets; Hilbert space; Poggio´s regularization theory; input space; input-output mapping; learning; neural nets; Boundary conditions; Computer science; Differential equations; Functional analysis; Laplace equations; Sufficient conditions;
Conference_Titel :
Neural Networks, 1993. IJCNN '93-Nagoya. Proceedings of 1993 International Joint Conference on
Print_ISBN :
0-7803-1421-2
DOI :
10.1109/IJCNN.1993.716985
