DocumentCode
147854
Title
A quantitative investment model based on multi-fractal theory and support vector machine
Author
Xudong Fan ; Hui Li ; Zhipu Zhu
Author_Institution
Shenzhen Grad. Sch., Shenzhen Eng. Lab. of Converged Networking Technol., Peking Univ., Shenzhen, China
fYear
2014
fDate
27-29 April 2014
Firstpage
239
Lastpage
244
Abstract
Quantitative investment which combines financial data with mathematics and computer technology has become a new rising method of the international investment community in recent ten years. This paper proposes a quantitative investment prediction model based on the latest mathematical theory and data analysis method. Firstly, we construct the whole quantitative investment system. Then we do qualitative analysis of financial market with multi-fractal method to see whether there exist fractal characteristics. Finally, we use support vector machine (SVM) to do quantitative analysis to predict changes in financial assets. We choose Shanghai Composite Index (000001.ss) as research target, test the model with five years of data and do error analysis on the output of the model. Our model can be used as quantitative investment strategies and is also useful for asset allocation in the future.
Keywords
fractals; investment; stock markets; support vector machines; SVM; Shanghai Composite Index; asset allocation; data analysis method; error analysis; financial asset change prediction; financial data; financial market; fractal characteristics; international investment community; mathematical theory; multifractal theory; quantitative investment prediction model; support vector machine; Algorithm design and analysis; Fractals; Indexes; Investment; Kernel; Stock markets; Support vector machines; SVM; financial market; multi-fractal; quantitative investment;
fLanguage
English
Publisher
ieee
Conference_Titel
Computing, Management and Telecommunications (ComManTel), 2014 International Conference on
Conference_Location
Da Nang
Print_ISBN
978-1-4799-2904-7
Type
conf
DOI
10.1109/ComManTel.2014.6825611
Filename
6825611
Link To Document