| Presentation | 2002/3/12 Utilization of a Modular-Type Neural Network with Nonlinear Transformation for Pattern Recognition Problems K. MORISHIMA, S. SATOH, M. SHIMOMURA, S. MIYAKE, H. ASO, |
|---|---|
| PDF Download Page | PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | A modular-type neural network is proposed in order to solve pattern recognition problems. Neurons in the hidden layer convert an input space to a curved space by use of polynominal functions, and a neuron in the output makes discriminant surfaces by RBF(Radial Basis Functions). The proposed network is able to recognize patterns whose distribution of the features are non-super-ellipsoids. We show that the network has ability to make a multiple discriminant surface, that is, multi-templates are generated automatically. Numeral simulations show that the network works successfully for different distributions of input: i) non-super-elliptic distribution, ii) concave distribution, iii) spatially separated distribution. We evaluate the network for The Monk's problem. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | modular-type neural network / RBF / nonlinear transformation / pattern recognition |
| Paper # | NC2001-180 |
| Date of Issue |
| Conference Information | |
| Committee | NC |
|---|---|
| Conference Date | 2002/3/12(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) | Utilization of a Modular-Type Neural Network with Nonlinear Transformation for Pattern Recognition Problems |
| Sub Title (in English) | |
| Keyword(1) | modular-type neural network |
| Keyword(2) | RBF |
| Keyword(3) | nonlinear transformation |
| Keyword(4) | pattern recognition |
| 1st Author's Name | K. MORISHIMA |
| 1st Author's Affiliation | Department of Electrical and Communication Engineering, Graduate School of Engineering,Tohoku University() |
| 2nd Author's Name | S. SATOH |
| 2nd Author's Affiliation | Department of Applied Physics, Graduate School of Engineering, Tohoku University |
| 3rd Author's Name | M. SHIMOMURA |
| 3rd Author's Affiliation | Department of Electrical and Communication Engineering, Graduate School of Engineering,Tohoku University |
| 4th Author's Name | S. MIYAKE |
| 4th Author's Affiliation | Department of Applied Physics, Graduate School of Engineering, Tohoku University |
| 5th Author's Name | H. ASO |
| 5th Author's Affiliation | Department of Electrical and Communication Engineering, Graduate School of Engineering,Tohoku University |
| Date | 2002/3/12 |
| Paper # | NC2001-180 |
| Volume (vol) | vol.101 |
| Number (no) | 736 |
| Page | pp.pp.- |
| #Pages | 8 |
| Date of Issue | |