資料探勘技術運用在股市序列樣型技術分析之研究

Applying Data Mining to Analyze Sequential Patterns in the stock

史育英(Yu-Ying Shih)

指導教授 ： 李鴻璋

淡江大學/管理學院/資訊管理學系碩士班/碩士(2006年)

https://doi.org/10.6846/TKU.2006.00777

摘要

本文利用資料探勘技術，來分析台灣股票期貨市場的研究，用過去股市的歷史Ｋ線資料建立序列樣型，來預測當日的漲跌，以提供金融投資決策者的參考。股票市場呈現所有投資者決策集合的結果，而投資者的投資行為有其時間關聯性；此投資順序性現象可以利用資料探勘(Data Mining)的技術進行知識的挖掘，尤其資料探勘中的時間序列樣型(Time Sequential Pattern)分析即是用來研究兩件事件前後順序的關係，而近來此項技術演算法有重大的突破，因此採用此一技術來分析股票市場的行為，希望可開展新的研究方向。
本研究的目的在於歸納台灣股票期貨市場的投資序列特徵，根據過去股票市場的歷史性Ｋ線交易資料來探勘出台灣股票期貨市場指數期貨的時間序列樣型相關模式，就是使用資料探勘的技術尋找期貨市場指數期貨中具有特別意義的Ｋ線組合序列，是否存在期貨市場開盤後15分鐘的三根五分鐘Ｋ線漲跌順序相關組合的特性，並依分析結果找出當天日Ｋ線具有預期漲跌幅度的相關特性，據此建構台灣股票期貨市場指數期貨的操作行為模式，以提供投資者作正確的金融投資決策。

關鍵字

資料探勘 ； 時間序列樣型 ； 股價預測

並列摘要

Abstract:
Our research applies data mining technique to analyze stock and forward market in Taiwan. We build the Time Sequential Pattern by constructing the historical data of Bar-chart (or K-line) of the stock market futures to assist investors` decisions making. The performance of the stock market is the collection of all individuals` decision, and there are some timing relations between their investments. The phenomenon of sequential investment can be studied or explained by using Data Mining technique, especially the Sequential Pattern Analysis. The Sequential Pattern Analysis is used to analyze the sequential relation between two events. The technique has advanced greatly in recent years, so we hope that it would be a new research way by using this technique to analyze the behavior of the stock market.
The object of this research is to generalize the characteristic of the time sequential pattern of investments in the stock market futures in Taiwan. Based on the historical bar-chart transaction data in the stock market, we mine for the sequential patterns of Taiwan stock exchange capitalization weighted stock index futures; use Data mining technique to discover some bar-chart combination sequence which may construct the behavior model to provide the investors with useful information and to assist them doing the correct decisions making .

並列關鍵字

Data Mining ； Time Sequential Pattern ； Price Forecasting

參考文獻

[9] 李安邦，「以遺傳演算法為基底的模糊專家系統於投資策略之應用」，元智管理研究所，碩士論文，1997。

[10] 林萍珍，「遺傳演算法在使用者導向的投資組合選擇之應用」，國立中央大學資管所，碩士論文，1998。

[22] A. Savasere, E. Omiecinski, and S. Navathe, “An Efficient Algorithm for Mining Association Rules,” Proceedings of the 21st Conference of Very Large Databases (VLDB), pp. 432-444 1995.

[23] C. Apte and S. Weiss. Data mining with decision trees and decision rules. Future Generation Computer Systems, 13, 1997.

[26] J. Han and M. Kamber, Data Mining: Concepts and Techniques, Morgan Kaufmann,San Francisco, 2001.