Responsive image
博碩士論文 etd-0923111-144703 詳細資訊
Title page for etd-0923111-144703
論文名稱
Title
型態辨識應用於技術分析指標之研究
Pattern Recognition of Technical Analysis Indicators
系所名稱
Department
畢業學年期
Year, semester
語文別
Language
學位類別
Degree
頁數
Number of pages
57
研究生
Author
指導教授
Advisor
召集委員
Convenor
口試委員
Advisory Committee
口試日期
Date of Exam
2011-06-18
繳交日期
Date of Submission
2011-09-23
關鍵字
Keywords
量化、編碼、交易策略、股價走勢、型態辨識、技術分析
trading strategy, pattern recognition, trend, technical analysis, encode
統計
Statistics
本論文已被瀏覽 5766 次,被下載 151
The thesis/dissertation has been browsed 5766 times, has been downloaded 151 times.
中文摘要
技術分析的基本概念建立在歷史將會不斷的重演之上,因此分析者能夠利用觀察過去的股價走勢來對未來的市場情勢進行推估,這樣的技術應用在近年來也更加廣泛。另一方面本文中的型態辨識技術的點子來自於臉部辨識的系統,在這樣的系統下,機器捕捉到被觀察者的臉部特徵並且利用編碼程序將所有的特徵量化形成一組足以描述目標形象的數據。透過比對這些數據的相似度,系統便可以確認被觀察者的身份。型態辨識應用了這樣的系統並且將編碼的想法轉移到了股價的走勢上,它將歷史的股價走勢高低點位等特徵利用量化的數字描述,藉此對未來的股價進行推測。在實際應用上,型態辨識可以作為一個交易策略的前製作業,搭配型態辨識出來的結果,交易策略也能夠進一步的修正出最有獲利能力的狀態。
這個模型取用的台灣加權指數作為分析的標的,總共提取了19個編碼,並可以區分為價格面向以及走勢面向。在經過比對以後就可以得出相似的預測結果。實證上利用了一個周頻的簡單策略這個模型達到了31.57%的年化報酬以及26.66%的年化風險。即便是在計入交易手續費以後,模型仍然具有14.94%的年化報酬以及26.72%的年化風險。
Abstract
In recent years technical analysis has been used more and more frequently. The original concept of technical analysis is built on history will be continue to repeat itself. Therefore, analysts and investors could predict the market price by observing the historical data.
The idea of pattern recognition technology comes from face recognition systems. In the system, the analyst captures the facial features from the entrant and then quantifies the features as codes. Through the process of recognition, the analyst can confirm the identity of the entrant. Pattern recognition applies the idea to extract information encoded in the stock market characteristics and recognize the market with historical data. In the application, pattern recognition can be regarded as a pre-operation of the technical analysis. Users analyze the current information through pattern recognition and can further build the strategy.
This model has 19 codes captured from two dimensions; the first is price, and the second is the trend of ups and downs. The empirical results for the decade in the weekly frequency trading strategy are an annual return of 31.57% and annual risk of 26.66%. After the deduction of trading fees, the strategy has an annual return of 14.94% and annual risk of 26.72%.
目次 Table of Contents
論文審定書 i
摘要 ii
Abstract iii
List of Tables vi
List of Figures vii
I. Introduction 1
1.1 Background 1
1.2 Research Motivation and Purpose 2
1.3 Research Framework and Flow 2
II. Literature Review 4
2.1 Technical analysis 4
2.2 Face recognition 6
III. Methodology 10
3.1 Feature Extractor 11
3.1.1 Data collection 11
3.1.2 Time span of the snapshot 11
3.1.3 Factor extraction 15
3.2 The PCA process 21
3.2.1 Principal component analysis (PCA) 21
3.2.2 Dimension Reduction 22
3.3 Pattern Recognition 23
3.3.1 Target period 23
3.3.2 Euclidean distance 23
3.3.3 Calculate the similar sample period 25
3.4 Application 27
IV. Empirical Results 28
4.1 Characteristic of recognition model 28
4.1.1 The distribution of Euclidean distance 28
4.1.2 Code – Type_6 28
4.2 Similar period trend graph 30
4.2.1 Database 30
4.2.2 Recognition trend graph (Target period: 20110422) 31
4.2.3 Recognition trend graph (Target period: 20090422) 35
4.2.4 Recognition trend graph (Target period: 20020422) 35
4.3 Back test of the rising probability 37
4.3.1 Statistic of future price with different number of sample periods 37
4.3.2 Frequency of database refresh 38
4.3.3 Estimation of target period 39
4.3.4 Accuracy of the recognition process 40
4.3.5 Empirical performance of recognition process 40
V. Conclusion 45
5.1 Conclusion 45
5.2 Suggestions for Further Research 46
References 48
參考文獻 References
Fama, Eugene (1970). Efficient Capital Markets: A Review of Theory and Empirical Work, Journal of Finance
Fama, Eugene (1965). The behavior of stock market prices, Journal of Business 38: 34–105.
Dreman D N, Berry M A. (1995). Analyst forecasting errors and implications for security analysis, Financial Analysts Journal, 1995, 51(3)
Pruitt, S. W., and White, R. E. (1988). The CRISMA Trading System: Who Says Technical Analysis Can not Beat The Market., Journal of Portfolio Management
Brock, Lakonishok and LeBaron (1992), Overall our results provide strong support for the technical strategies that we explored
Alexander, S. S. (1961) Price movement in speculative markets: trends or
random walks. Industrial Management Review
Levy, R. A. (1967). Relative strength as a criterion for investment selection.
Journal of Finance
Jensen, M. C. and Benington, G. A. (1970). Random walks and technical
theories: some additional evidence. Journal of Finance
Bohan, James (1981), Relative Strength: Further Positive Evidence. Journal of Portfolio Management
Cootner, Paul H. (1964). The random character of stock market prices. MIT Press.
Van Horne, J. C. and G. G. C. Parker (1967), The random walk theory: An empirical test, Financial Analysts Journal
M. Turk, A. Pentland(1991). Eigenfaces for Recognition, Journal of Cognitive Neurosicence
L. D. Harmon, M. K. Khan, R. Lasch, and P. F. Ramig (1981). Machine identification of human faces, Pattern Recognition
A. J. Goldstein, L. Harmon, and A. Lesk, Identification of human faces, Proceedings of the TEEE
Hsun-Li Chang(2002). Face Recognition Using Principal Facial-Factor Analysis, Datung university
Hyeonjoon Moon and Jonathon Phillips (2001). Computational and performance aspects of PCA-based face-recognition algorithm
Chen-chieh Wang (2005),technical analysis, trading rule, SPA test, modified simulation trade Choyang university
電子全文 Fulltext
本電子全文僅授權使用者為學術研究之目的,進行個人非營利性質之檢索、閱讀、列印。請遵守中華民國著作權法之相關規定,切勿任意重製、散佈、改作、轉貼、播送,以免觸法。
論文使用權限 Thesis access permission:自定論文開放時間 user define
開放時間 Available:
校內 Campus: 已公開 available
校外 Off-campus: 已公開 available


紙本論文 Printed copies
紙本論文的公開資訊在102學年度以後相對較為完整。如果需要查詢101學年度以前的紙本論文公開資訊,請聯繫圖資處紙本論文服務櫃台。如有不便之處敬請見諒。
開放時間 available 已公開 available

QR Code