A rule extraction based approach in predicting derivative use for financial risk hedging by construction companies

NCUIR > Engineering College > Graduate Institute of construction engineering > journal & Dissertation > Item 987654321/50071

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/50071

Title:	A rule extraction based approach in predicting derivative use for financial risk hedging by construction companies
Authors:	Chen,JH;Yang,LR;Su,MC;Lin,JZ
Contributors:	營建管理研究所
Keywords:	FUZZY;MODEL
Date:	2010
Issue Date:	2012-03-27 17:02:49 (UTC+8)
Publisher:	國立中央大學
Abstract:	Prevention of financial risk is one of the major tasks that construction companies have to pay attention to. Using derivatives to avoid such risks is a practical strategy, but is heavily dependent on the traders' skills and accuracy of predictions. The purpose of this study is to develop an automatic expert model using a rule extraction based approach that provides practitioners with a prediction tool for the hedging of financial risks through the use of derivatives. Data for the study include 780 quarterly financial statements collected from 2002 to 2006, based on public information from 39 listed construction companies in Taiwan. Statements with incomplete and missing data are eliminated, leaving 672 with which to construct the rule extraction based model, the Hyper Rectangular Composite Neural Networks (HRCNNs). After factor dimension reduction, only 16 financial ratios out of all revealed ratios are left to be used as input variables. The HRCNNs yield an 80.6% successful classification rate. With these 16 financial ratios and the proposed model, derivative use to hedge financial risk can be established for the benefit of the construction practitioners. (C) 2010 Elsevier Ltd. All rights reserved.
Relation:	EXPERT SYSTEMS WITH APPLICATIONS
Appears in Collections:	[Graduate Institute of construction engineering] journal & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	726	View/Open

社群 sharing

Loading...