基於交易資訊及市場訊息的全面性特徵於智慧型股市交易模型之研究

An Intelligent Stock Trading Model from Comprehensive Features Based on Trading Information and Market Messages

吳政隆(Jheng-Long Wu)

指導教授 ： 張百棧

元智大學/資訊學院/資訊管理學系/博士(2013年)

https://doi.org/10.6838/YZU.2013.00075

摘要

本論文主要目的是預測股票市場的自動化交易決策。股票市場是一種相當複雜的交易環境，對於全面性特徵的產生是非常困難。根據許多研究結果指出，投資者掌握相關股市資訊可對投資標的物獲取更多利潤。本研究嘗試發展一套基於全面性特徵的智慧型股市交易模型(Intelligent Stock Trading Model Based on Comprehensive features, ISTMCF)來預測未來的股票趨勢，特徵包括情感指標(Sentiment indices, SI)、技術指標(Technical indices, TI)、交易訊號(Trading signals, TS)。此ISTMCF模式的組成包含以股市資訊萃取、預測模式學習及股市交易決策。全面性特徵的萃取方法包含，情感分析(Sentiment analysis, SA)可即時分析較為敏感的事件訊息有助於推估未來趨勢形成情感指標(Sentiment indices, SI)，而技術分析(Technical analysis, TA)可以從歷史交易資訊中計算有效的交易規則形成技術指標(Technical indices, TI)，然而趨勢切割法(Trend-based segmentation method, TBSM)可從歷史股價中計算出更為貼切現實的交易決策形成交易訊號(Trading signals, TS)。相較於其他未產用情感指標為基礎的交易模型，本研究所提出的全面性特徵可大幅提升股市預測的精準度及獲利率。

關鍵字

智慧型股票交易模型 ； 全面性特徵 ； 情感分析 ； 技術分析 ； 機器學習

並列摘要

The aim of this study is to predict automatic trading decisions in the stock market. Comprehensive feature for predicting future trend is very difficult to generate in a complex environment. According to related works, the most relevant stock information can help investors formulate objects that will result in better profit. With this in mind, we present a model of intelligent stock trading model from comprehensive features (ISTMCF) to predict future stock market trend. The ISTMCF consists of stock information extraction, prediction model learning and stock trading decision. We apply three approaches to generate comprehensive features, including sentiment analysis (SA) that provides sensitive market events from news on stock for sentiment indices (SI), technical analysis (TA) that yields effective trading rules based on trading information on the stock exchange for technical indices (TI), as well as trend-based segmentation method (TBSM) that raises trading decisions from stock price for trading signals TS). Through experiment, we demonstrate that for all stocks, the results of employing the comprehensive features are significantly better than that without using the sentiment analysis features.

並列關鍵字

Intelligent Stock Trading Model ； Comprehensive Features ； Sentiment Analysis ； Technical Analysis ； Machine Learning

參考文獻

[1]. T. Martikainen, J. Perttunen, P. Yli-Olli, and A. Gunasekaran, 1994. The impact of the return interval on common factors in stock returns: Evidence from a thin security market, Journal of Banking & Finance, 18(4), 659-672.

[2]. D. Kirkpatrick, and R. Dahlquist, 2006. Technical analysis: The complete resource for financial market technicians, Financial Times Press, 2006, page 3.

[3]. P. C. Chang, and C. Y. Fan, 2008. A hybrid system integrating a wavelet and TSK fuzzy rules for stock price forecasting, IEEE Transactions on Systems, Man, and Cybernetics – Part C: Applications and Reviews, 38(6), 802-815.

[4]. P. C. Chang, T. W. Liao, J. J. Lin, and C. Y. Fan, 2011. A dynamic threshold decision system for stock trading signal detection, Applied Soft Computing, 11(5), 3998-4010.

[5]. S. Asadi, E. Hadavandi, F. Mehmanpazir, and M. M. Nakhostin, 2012. Hybridization of evolutionary Levenberg–Marquardt neural networks and data pre-processing for stock market prediction, Knowledge-Based Systems, 35(11), 245-258.