Using Artificial Intelligence In Prediction Of Taiwan Stock Markets

碩士 === 國立中山大學 === 資訊管理研究所 === 85 ===   This thesis includes two parts, one is comp‘arative research: using literature review techniques to compare the difference between the systems proposed in the literature. The other part construct an intelligent stock decision support system to support stock inv...

Full description

Bibliographic Details
Main Authors: Hsieh, Ching-Hao, 謝青皓
Other Authors: Liang, Ting-Peng
Format: Others
Language:zh-TW
Published: 1997
Online Access:http://ndltd.ncl.edu.tw/handle/37174506394971499196
Description
Summary:碩士 === 國立中山大學 === 資訊管理研究所 === 85 ===   This thesis includes two parts, one is comp‘arative research: using literature review techniques to compare the difference between the systems proposed in the literature. The other part construct an intelligent stock decision support system to support stock investment.   There are two major components in this system. The first one is artificial neural networks system. It can generate next day’s trading signal. The other one is an expert system, consists of many trading rules generated by using rule induction to induce the rules of different technical indicators.   The performance of the system is evaluated by trading the Taiwan Stock Exchange Weighted Price Index from 1994/01 to 1997/02. The annual rates of returns by the system is not only higher than those suing buy-and-hold strategy, but also superior to all the closed-end funds in the past three years.