Doctoral Dissertations
Abstract
"This research presents a study of intelligent stock price forecasting systems using interval type-2 fuzzy logic for analyzing Japanese candlestick techniques. Many intelligent financial forecasting models have been developed to predict stock prices, but many of them do not perform well under unstable market conditions. One reason for poor performance is that stock price forecasting is very complex, and many factors are involved in stock price movement. In this environment, two kinds of information exist, including quantitative data, such as actual stock prices, and qualitative data, such as stock traders' opinions and expertise. Japanese candlestick techniques have been proven to be effective methods for describing the market psychology. This study is motivated by the challenges of implementing Japanese candlestick techniques to computational intelligent systems to forecast stock prices. The quantitative information, Japanese candlestick definitions, is managed by type-2 fuzzy logic systems. The qualitative data sets for the stock market are handled by a hybrid type of dynamic committee machine architecture. Inside this committee machine, generalized regression neural network-based experts handle actual stock prices for monitoring price movements. Neural network architecture is an effective tool for function approximation problems such as forecasting. Few studies have explored integrating intelligent systems and Japanese candlestick methods for stock price forecasting. The proposed model shows promising results. This research, derived from the interval type-2 fuzzy logic system, contributes to the understanding of Japanese candlestick techniques and becomes a potential resource for future financial market forecasting studies"--Abstract, page iii.
Advisor(s)
Dagli, Cihan H., 1949-
Committee Member(s)
Gelles, Gregory M.
Long, Suzanna, 1961-
Cudney, Elizabeth A.
Allada, Venkat
Department(s)
Engineering Management and Systems Engineering
Degree Name
Ph. D. in Engineering Management
Publisher
Missouri University of Science and Technology
Publication Date
Fall 2011
Pagination
xiii, 155 pages
Note about bibliography
Includes bibliographical references (pages 140-154).
Rights
© 2011 Takenori Kamo, All rights reserved.
Document Type
Dissertation - Open Access
File Type
text
Language
English
Subject Headings
Economic forecasting -- Mathematical modelsEconomic forecasting -- MethodologyFuzzy logicInvestment analysis -- Mathematical modelsStocks -- Charts, diagrams, etc
Thesis Number
T 9897
Print OCLC #
794749233
Electronic OCLC #
747978286
Recommended Citation
Kamo, Takenori, "Integrated computational intelligence and Japanese candlestick method for short-term financial forecasting" (2011). Doctoral Dissertations. 1908.
https://scholarsmine.mst.edu/doctoral_dissertations/1908