Paper published in a book (Scientific congresses and symposiums)
Automatic Learning of General Type-2 Fuzzy Logic Systems using Simulated Annealing
Almaraashi, Majid; John, Robert; Hopgood, Adrian
2014In 2014 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2014)
Peer reviewed
 

Files


Full Text
fuzz-ieee2014.pdf
Publisher postprint (444.21 kB)
Request a copy

All documents in ORBi are protected by a user license.

Send to



Details



Abstract :
[en] This paper reports on a new approach for automatic learning of general type-2 fuzzy logic systems (GT2FLSs) using simulated annealing (SA). The learning process in this work starts without an initial interval type-2 fuzzy system and has an objective to optimize all membership function parameters involved in the general type-2 fuzzy set in two stages. This is a novel methodology for learning GT2FLSs using the vertical-slices representation. The methodology used here is based on a proposed parameterization method presented in a previous work to ease the design of GT2FLSs. Two models of GT2FLSs have been applied using two different type-reduction techniques. The first technique is the sampling method, which is non-deterministic. The second technique is the vertical-slices centroid type-reduction (VSCTR), which is deterministic. Both models as well as an interval type-2 fuzzy logic system (IT2FLS) model have been applied to predict a Mackey-Glass time series. A comparison of the results of modeling these problems using the three models showed more accurate modeling for the GT2FLSs when using the VSCTR deterministic defuzzification method. It has also been shown that a GT2FLS with VSCTR defuzzification is more able to handle uncertainty than an IT2FLS, although the latter was faster.
Disciplines :
Computer science
Electrical & electronics engineering
Author, co-author :
Almaraashi, Majid;  Umm Al Qura Univ, Univ Coll Aljamoum, Mecca, Saudi Arabia.
John, Robert;  Univ Nottingham, Automated Scheduling Optimisat & Planning Grp ASA, Nottingham NG7 2RD, England.
Hopgood, Adrian ;  Université de Liège > HEC - Ecole de gestion de l'ULG : Direction générale
Language :
English
Title :
Automatic Learning of General Type-2 Fuzzy Logic Systems using Simulated Annealing
Publication date :
2014
Event name :
IEEE International Conference on Fuzzy Systems
Event organizer :
IEEE
Event place :
Beijing, China
Event date :
JUL 06-11, 2014
Audience :
International
Main work title :
2014 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2014)
Publisher :
IEEE
ISBN/EAN :
9781479920747
Pages :
2384-2390
Peer reviewed :
Peer reviewed
Available on ORBi :
since 11 February 2016

Statistics


Number of views
17 (1 by ULiège)
Number of downloads
0 (0 by ULiège)

Scopus citations®
 
1
Scopus citations®
without self-citations
0

Bibliography


Similar publications



Contact ORBi