以綜合式人工類神經網路方法 預測股價指數:以泰國為例

CCUR > College of Business > Department of Business Administration & Graduate Institute of International Business Administration > Thesis > Item 987654321/28335

Please use this identifier to cite or link to this item: https://irlib.pccu.edu.tw/handle/987654321/28335

Title:	以綜合式人工類神經網路方法預測股價指數:以泰國為例 Forecasting stock index with an integrated ANN method: A Thailand case
Authors:	周若瑄 Kamsorn, Parichat
Contributors:	國際企業管理學系
Keywords:	forecasting stock price artificial neural networks backpropagation wavelet neural network SET index
Date:	2014-06
Issue Date:	2014-09-30 17:14:11 (UTC+8)
Abstract:	Stock price is basically sensitive, non-stationary and very noisy. Many environmental factors are the important variables in stock price change, especially in emerging markets. To forecast stock price, this study proposed the developed integration artificial neural networks (ANNs) using the Wavelet De-nosing-based Back propogation (WDBP) neural network. The main purpose of the wavelet de-composition is to classify the basic elements from the noise of the signal. The used data in this experiment were the monthly closing prices of Stock Exchange of Thailand (SET) index during January 2001 to April 2014. To show the improved integration of using WDBP method, this paper applied three accurate measures to evaluate the forecasting performance. Following this paper methodology, the investors could be guided in investment providing deviation and direction of stock indexes and maximization profits in the emerging stock market.
Appears in Collections:	[Department of Business Administration & Graduate Institute of International Business Administration ] Thesis

Files in This Item:

File	Description	Size	Format
fb140930171348.pdf		2017Kb	Adobe PDF	532	View/Open

View Licence