Time series prediction using crisp and fuzzy neural networks: a comparative study
Dissertant
Thesis advisor
Comitee Members
Smith, Kevin
Driouchi, Ahmad
Palliam, Ralph
University
Al Akhawayn University
Faculty
School of Science and Engineering
Department
Computer Science
University Country
Morocco
Degree
Master
Degree Date
1999
English Abstract
Every organization needs adequate forecasts for planning the future.
The accuracy of forecasts is influenced by both the quality of the past data and the method selected to forecast the future.
In this thesis, we have made a comparative study between the time series forecasts from conventional neural networks (quick-propagation), fuzzy-neural networks (Adaptive-Network-based Fuzzy Inference System (ANFIS)) and those from traditional time series methods (regression and ARIMA models).
We use fuzzy curves to identify the input variables that have most influence on the output.
This method identifies the significant input variables that lead to considerable decrease in training time for ANFIS.
We test the performance of quick-propagation and ANFIS with the fuzzy curve pruning technique on different empirical time series data (the Gross Domestic Product and the National Private Consumption), the Zimmermann and Zysno data and Monte Carlo simulated time series production data.
The performance of ANFIS is superior to that of the quick-propagation, regression and ARIMA models on the empirical time series data (the Gross Domestic Product and the National Private Consumption) and the Zimmermann and Zysno data.
On the other hand, ANFIS does not perform as well on the time series production data, where ANFIS fails to leam the right input fuzzy sets from the training data.
Main Subjects
Information Technology and Computer Science
Topics
No. of Pages
87
Table of Contents
Table of contents.
Abstract.
Abstract in Arabic.
Chapter One : Introduction.
Chapter Two : Neural networks and fuzzy logic in time series forecasting.
Chapter Three : Time series models in economy.
Chapter Four : Experiments and results.
Chapter Five : Conclusion and future work.
References.
American Psychological Association (APA)
Bouqata, Bushra. (1999). Time series prediction using crisp and fuzzy neural networks: a comparative study. (Master's theses Theses and Dissertations Master). Al Akhawayn University, Morocco
https://search.emarefa.net/detail/BIM-629977
Modern Language Association (MLA)
Bouqata, Bushra. Time series prediction using crisp and fuzzy neural networks: a comparative study. (Master's theses Theses and Dissertations Master). Al Akhawayn University. (1999).
https://search.emarefa.net/detail/BIM-629977
American Medical Association (AMA)
Bouqata, Bushra. (1999). Time series prediction using crisp and fuzzy neural networks: a comparative study. (Master's theses Theses and Dissertations Master). Al Akhawayn University, Morocco
https://search.emarefa.net/detail/BIM-629977
Language
English
Data Type
Arab Theses
Record ID
BIM-629977
