Neurofuzzy characterization of financial time series in an anticipatory framework

Author

Pantazopoulos, K.N. ; Tsoukalas, L.H. ; Houstis, E.N.

Author_Institution

Purdue Univ., West Lafayette, IN, USA

fYear

1997

fDate

23-25 Mar 1997

Firstpage

50

Lastpage

56

Abstract

Neurofuzzy characterization of financial time series refers to the judicious application of neural and fuzzy tools to the problem of time series prediction. A methodology is presented where fuzzy if/then rules and neural predictors are used to anticipate the predictability a time series over various time horizons. The methodology is tested with actual financial time series data (S&S 500 daily closes) and shows considerable promise as a decision making and planning tool. Results in the context of option trading strategies are presented and discussed

Keywords

decision support systems; financial data processing; fuzzy logic; neural nets; planning (artificial intelligence); prediction theory; time series; anticipatory framework; decision making tool; financial time series; fuzzy if/then rules; fuzzy tools; neural predictors; neural tools; neurofuzzy characterization; option trading strategies; planning tool; predictability; time horizons; time series prediction; Computer architecture; Context modeling; Decision making; Degradation; Economic forecasting; Fuzzy logic; Neural networks; Predictive models; Testing; Time measurement;

fLanguage

English

Publisher

ieee

Conference_Titel

Computational Intelligence for Financial Engineering (CIFEr), 1997., Proceedings of the IEEE/IAFE 1997

Conference_Location

New York City, NY

Print_ISBN

0-7803-4133-3

Type

conf

DOI

10.1109/CIFER.1997.618904

Filename

618904