The Study of Comparison in Trading Decision from Forecasting StockPrice By Artificial Neural Network and Technical Analysis

• Main Authors: Chi-Ting Lai, 賴季廷
• Other Authors: none
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/18968828435540068112

碩士 === 國立雲林科技大學 === 工業工程與管理研究所碩士班 === 99 === 20 objective stocks are the greatest weight in every industry index from 20 industry indexes were established by TWSE in 2006. We can import factor of risk to model Classification and Regression Tree （CART） and grouping. 15 objective stocks is a high hazard group and 5 objective stocks is a low hazard group. We can import 12 daily data to Artificial Neural Network （ANN） and forecasting stock price. Making trading decision by continual signal with the forecasting stock price and computing investment return. Making trading decision by technical indicators and computing investment return. The result is that making trading decision by continual signal is better than technical indicators. The technical indicators are failure when the stock price is over and over raise. It is no use making trading decision and getting lower investment return. The continual signal trading decision was not restricted in this situation and we make correct decision to obtain better investment return.