運用遺傳基因演算法建立動態證券市場技術模型

Abstract

本研究的主要目的是希望結合非線性股市預測與基因演算法的空間搜尋能 力，提出一個股市非線性的預測模型，並且實際以電腦擬模擬本研究模型 並驗證模型的預測能力。本研究主要分為三部份，第一部份是股市非線性 預測模型的設計與定義；第二部份是基因演算法空間搜尋方式的設計；第 三部份模型的實作與測試。最後，這個實作發現本研究的模型確實能夠預 測非結構化的股市。未來，本研究尚可考量基本面指標與技術面指標的整 合運用模型上，以期得到更佳的預測結果。 The Propose of this research is to combine the non-linear stock price forecasting and the space searching capacity of the Genetic algorithm ,and to propose a non-linear stock market forecasting model,and to to certify model's forcasting capacit. There are three part of this research . First is the design and definition of the non-linear stock forecasting model;Second is the design of the genetic algorithm searching architecture. Third is the implement and testing of the model. Finally,the research In the future,the research mode can also consider the integration of the economical index and technical index in order to get a more precise result.