Spatial interpolation using MLP-RBFN hybrid networks

I. Cheng Yeh, Kuan Chieh Huang, Yau-Hwang Kuo

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

It is easy for a multi-layered perception (MLP) to fit a stratified spatial interpolation pattern whose form is close to open surface; while it is easy for a radial basis function network (RBFN) to fit a pocket (radial) spatial interpolation pattern whose form is close to closed surface. However, in the real world, the spatial interpolation pattern may consist of stratified and pocket patterns. Neither MLP nor RBFN can fit the pattern easily. To combine their advantages to fit the complex hybrid spatial interpolation patterns, in this article we propose a novel neural network, MLP-RBFN hybrid network (MRHN), whose hidden layer contains sigmoid and Gaussian units at the same time. Although there are two kinds of processing units in MRHN, in this study we used the principle of minimizing the error sum of squares to derive the supervised learning rules for all the network parameters. This research took rainfall distribution in Taiwan as a case study. The results show that (1) the prediction error of the testing dataset outside the training dataset demonstrated that MRHN was the most accurate among the three networks, RBFN was the next best, and MLP was the worst; (2) the MLP model seriously underestimated the values of high observed rainfall; (3) over-learning may be a serious shortcoming of using RBFN in spatial interpolation applications; (4) MRHN may have better generalization learning capacity than RBFN in spatial interpolation applications.

Original languageEnglish
Pages (from-to)1884-1901
Number of pages18
JournalInternational Journal of Geographical Information Science
Volume27
Issue number10
DOIs
Publication statusPublished - 2013 Oct 1

Fingerprint

Radial basis function networks
interpolation
Interpolation
Rain
learning
rainfall
Supervised learning
Neural networks
Testing
Processing
prediction
neural network
Taiwan

All Science Journal Classification (ASJC) codes

  • Information Systems
  • Geography, Planning and Development
  • Library and Information Sciences

Cite this

@article{1f2fcd17470949d29511433bd578cf5c,
title = "Spatial interpolation using MLP-RBFN hybrid networks",
abstract = "It is easy for a multi-layered perception (MLP) to fit a stratified spatial interpolation pattern whose form is close to open surface; while it is easy for a radial basis function network (RBFN) to fit a pocket (radial) spatial interpolation pattern whose form is close to closed surface. However, in the real world, the spatial interpolation pattern may consist of stratified and pocket patterns. Neither MLP nor RBFN can fit the pattern easily. To combine their advantages to fit the complex hybrid spatial interpolation patterns, in this article we propose a novel neural network, MLP-RBFN hybrid network (MRHN), whose hidden layer contains sigmoid and Gaussian units at the same time. Although there are two kinds of processing units in MRHN, in this study we used the principle of minimizing the error sum of squares to derive the supervised learning rules for all the network parameters. This research took rainfall distribution in Taiwan as a case study. The results show that (1) the prediction error of the testing dataset outside the training dataset demonstrated that MRHN was the most accurate among the three networks, RBFN was the next best, and MLP was the worst; (2) the MLP model seriously underestimated the values of high observed rainfall; (3) over-learning may be a serious shortcoming of using RBFN in spatial interpolation applications; (4) MRHN may have better generalization learning capacity than RBFN in spatial interpolation applications.",
author = "Yeh, {I. Cheng} and Huang, {Kuan Chieh} and Yau-Hwang Kuo",
year = "2013",
month = "10",
day = "1",
doi = "10.1080/13658816.2013.769050",
language = "English",
volume = "27",
pages = "1884--1901",
journal = "International Journal of Geographical Information Science",
issn = "1365-8816",
publisher = "Taylor and Francis Ltd.",
number = "10",

}

Spatial interpolation using MLP-RBFN hybrid networks. / Yeh, I. Cheng; Huang, Kuan Chieh; Kuo, Yau-Hwang.

In: International Journal of Geographical Information Science, Vol. 27, No. 10, 01.10.2013, p. 1884-1901.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Spatial interpolation using MLP-RBFN hybrid networks

AU - Yeh, I. Cheng

AU - Huang, Kuan Chieh

AU - Kuo, Yau-Hwang

PY - 2013/10/1

Y1 - 2013/10/1

N2 - It is easy for a multi-layered perception (MLP) to fit a stratified spatial interpolation pattern whose form is close to open surface; while it is easy for a radial basis function network (RBFN) to fit a pocket (radial) spatial interpolation pattern whose form is close to closed surface. However, in the real world, the spatial interpolation pattern may consist of stratified and pocket patterns. Neither MLP nor RBFN can fit the pattern easily. To combine their advantages to fit the complex hybrid spatial interpolation patterns, in this article we propose a novel neural network, MLP-RBFN hybrid network (MRHN), whose hidden layer contains sigmoid and Gaussian units at the same time. Although there are two kinds of processing units in MRHN, in this study we used the principle of minimizing the error sum of squares to derive the supervised learning rules for all the network parameters. This research took rainfall distribution in Taiwan as a case study. The results show that (1) the prediction error of the testing dataset outside the training dataset demonstrated that MRHN was the most accurate among the three networks, RBFN was the next best, and MLP was the worst; (2) the MLP model seriously underestimated the values of high observed rainfall; (3) over-learning may be a serious shortcoming of using RBFN in spatial interpolation applications; (4) MRHN may have better generalization learning capacity than RBFN in spatial interpolation applications.

AB - It is easy for a multi-layered perception (MLP) to fit a stratified spatial interpolation pattern whose form is close to open surface; while it is easy for a radial basis function network (RBFN) to fit a pocket (radial) spatial interpolation pattern whose form is close to closed surface. However, in the real world, the spatial interpolation pattern may consist of stratified and pocket patterns. Neither MLP nor RBFN can fit the pattern easily. To combine their advantages to fit the complex hybrid spatial interpolation patterns, in this article we propose a novel neural network, MLP-RBFN hybrid network (MRHN), whose hidden layer contains sigmoid and Gaussian units at the same time. Although there are two kinds of processing units in MRHN, in this study we used the principle of minimizing the error sum of squares to derive the supervised learning rules for all the network parameters. This research took rainfall distribution in Taiwan as a case study. The results show that (1) the prediction error of the testing dataset outside the training dataset demonstrated that MRHN was the most accurate among the three networks, RBFN was the next best, and MLP was the worst; (2) the MLP model seriously underestimated the values of high observed rainfall; (3) over-learning may be a serious shortcoming of using RBFN in spatial interpolation applications; (4) MRHN may have better generalization learning capacity than RBFN in spatial interpolation applications.

UR - http://www.scopus.com/inward/record.url?scp=84885947686&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84885947686&partnerID=8YFLogxK

U2 - 10.1080/13658816.2013.769050

DO - 10.1080/13658816.2013.769050

M3 - Article

VL - 27

SP - 1884

EP - 1901

JO - International Journal of Geographical Information Science

JF - International Journal of Geographical Information Science

SN - 1365-8816

IS - 10

ER -