ARIMA-BPN時間數列神經網路

CHUR > College Computer Science and Informatics > Department of Information Management > Journal Articles > Item 987654321/32037

Please use this identifier to cite or link to this item: http://chur.chu.edu.tw/handle/987654321/32037

Title:	ARIMA-BPN時間數列神經網路
Authors:	葉怡成 Yeh, I-Cheng
Contributors:	資訊管理學系 Information Management
Keywords:	時間數列;ARIMA;倒傳遞神經網路 time;Network
Date:	2009
Issue Date:	2014-06-27 01:42:19 (UTC+8)
Abstract:	本研究提出一個結合ARIMA與倒傳遞網路(Back-Propagation Network, BPN)優點的ARIMA-BPN神經網路，它是以BPN為模型，將ARIMA模式的輸入，包括前時p個時刻的數列值與前q個時刻的數列殘差值做為輸入值，組成的非線性函數，以建立更準確的時間數列預測模型。因為數列殘差值在BPN的訓練過程中會因網路連結權值的調整而改變，因此必須修改BPN的演算法來適應此需求，即藉由不斷更新每次預測所得之殘差值做為網路的輸入值。本研究以六個人為設計的例題，及四個現實世界的例題來比較ARIM In this paper we propose an ARIMA-BPN algorithm combining the advantages of ARIMA and Back-propagation networks (BPN). The algorithm is based on BPN and its inputs are the same as ARIMA. It can generate a non-linear function to create an accurate model
Appears in Collections:	[Department of Information Management] Journal Articles

Files in This Item:

File	Description	Size	Format
p_m321_0019.pdf		33Kb	Adobe PDF	243	View/Open