| Presentation | 1993/12/14 Regularization Networks and Learning Algorithms for Approximating Multi-Valued Functions Masahiko Shizawa, |
|---|---|
| PDF Download Page | PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | The regularization network(RN)is extended to approximate multi- valued functions so that one-to-h mapping,where h denotes multiplicity of the mapping,can be represented and learned from a finite number of inputoutput samples without clustering operations on the sample data set.Multi-valued function approximations are useful for learning ambiguous input-output relations from examples. This extension,which we call the Multi-Valued Regularization Network(MVRN),is derived from the Multi-Valued Standard Regularization Theory(MVSRT),which is an extension of standard regularization theory to multi-valued functions.MVSRT is based on a direct algebraic representation of multi-valued functions.By simple transformation of the unknown functions,we can obtain linear Euler-Lagrange equations.Therefore,the learning algorithm for MVRN is reduced to solving a linear system.It′s rather surpris ing that the dimension of the linear system is invariant to the multiplicity h.The proposed theory can be specialized and extended into Radial Basis Function(RBF)Methods,Generalized RBF(GRBF),and HyperBF networks of multi-valued functions.We also describe how the vector-valued function approximations can be extended into the multi-and vector-valued function approximations. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Regularization Network / Standard Regularization Theory / Multi- Valued Functions / One-to-many mapping / Computational learning / Feedforward neural network |
| Paper # | NC93-66 |
| Date of Issue |
| Conference Information | |
| Committee | NC |
|---|---|
| Conference Date | 1993/12/14(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | |
| Vice Chair | |
| Secretary | |
| Assistant | |
| Paper Information | |
| Registration To | Neurocomputing (NC) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Regularization Networks and Learning Algorithms for Approximating Multi-Valued Functions |
| Sub Title (in English) | |
| Keyword(1) | Regularization Network |
| Keyword(2) | Standard Regularization Theory |
| Keyword(3) | Multi- Valued Functions |
| Keyword(4) | One-to-many mapping |
| Keyword(5) | Computational learning |
| Keyword(6) | Feedforward neural network |
| 1st Author's Name | Masahiko Shizawa |
| 1st Author's Affiliation | ATR Human Information Processing Research Laboratories() |
| Date | 1993/12/14 |
| Paper # | NC93-66 |
| Volume (vol) | vol.93 |
| Number (no) | 376 |
| Page | pp.pp.- |
| #Pages | 8 |
| Date of Issue | |