Equivalence between Weight Decay Learning and Explicit Regularization to Improve Fault Tolerance of RBF

碩士 === 朝陽科技大學 === 網路與通訊研究所 === 96 === This thesis studies the fault tolerant performance of a radial basis function (RBF) network being trained to against the multiplicative weight noise (MWN) using both weight decay learning (WDL) and explicit regularization learning (ERL). It is shown analytical t...

Full description

Bibliographic Details
Main Authors:	Wun-He Luo, 羅文和
Other Authors:	John Sum
Format:	Others
Language:	zh-TW
Published:	2008
Online Access:	http://ndltd.ncl.edu.tw/handle/91062868963543008427

id	ndltd-TW-096CYUT5650006
record_format	oai_dc
spelling	ndltd-TW-096CYUT56500062015-11-27T04:04:14Z http://ndltd.ncl.edu.tw/handle/91062868963543008427 Equivalence between Weight Decay Learning and Explicit Regularization to Improve Fault Tolerance of RBF 權重衰退學習與明確正規化學習於幅狀基底函數網路之容錯訓練 Wun-He Luo 羅文和碩士朝陽科技大學網路與通訊研究所 96 This thesis studies the fault tolerant performance of a radial basis function (RBF) network being trained to against the multiplicative weight noise (MWN) using both weight decay learning (WDL) and explicit regularization learning (ERL). It is shown analytical that the fault tolerant abilities of WDL and ERL against MWN are similar if the number of training data is large and is demonstrated by simulations on 1D and 2D problems. John Sum Yung-Fa Huang 沈培輝黃永發 2008 學位論文 ; thesis 59 zh-TW
collection	NDLTD
language	zh-TW
format	Others
sources	NDLTD
description	碩士 === 朝陽科技大學 === 網路與通訊研究所 === 96 === This thesis studies the fault tolerant performance of a radial basis function (RBF) network being trained to against the multiplicative weight noise (MWN) using both weight decay learning (WDL) and explicit regularization learning (ERL). It is shown analytical that the fault tolerant abilities of WDL and ERL against MWN are similar if the number of training data is large and is demonstrated by simulations on 1D and 2D problems.
author2	John Sum
author_facet	John Sum Wun-He Luo 羅文和
author	Wun-He Luo 羅文和
spellingShingle	Wun-He Luo 羅文和 Equivalence between Weight Decay Learning and Explicit Regularization to Improve Fault Tolerance of RBF
author_sort	Wun-He Luo
title	Equivalence between Weight Decay Learning and Explicit Regularization to Improve Fault Tolerance of RBF
title_short	Equivalence between Weight Decay Learning and Explicit Regularization to Improve Fault Tolerance of RBF
title_full	Equivalence between Weight Decay Learning and Explicit Regularization to Improve Fault Tolerance of RBF
title_fullStr	Equivalence between Weight Decay Learning and Explicit Regularization to Improve Fault Tolerance of RBF
title_full_unstemmed	Equivalence between Weight Decay Learning and Explicit Regularization to Improve Fault Tolerance of RBF
title_sort	equivalence between weight decay learning and explicit regularization to improve fault tolerance of rbf
publishDate	2008
url	http://ndltd.ncl.edu.tw/handle/91062868963543008427
work_keys_str_mv	AT wunheluo equivalencebetweenweightdecaylearningandexplicitregularizationtoimprovefaulttoleranceofrbf AT luówénhé equivalencebetweenweightdecaylearningandexplicitregularizationtoimprovefaulttoleranceofrbf AT wunheluo quánzhòngshuāituìxuéxíyǔmíngquèzhèngguīhuàxuéxíyúfúzhuàngjīdǐhánshùwǎnglùzhīróngcuòxùnliàn AT luówénhé quánzhòngshuāituìxuéxíyǔmíngquèzhèngguīhuàxuéxíyúfúzhuàngjīdǐhánshùwǎnglùzhīróngcuòxùnliàn
_version_	1718137350919290880

Equivalence between Weight Decay Learning and Explicit Regularization to Improve Fault Tolerance of RBF

Similar Items