Decomposing Conventional Fuzzy Logic Systems to Hierarchical Fuzzy Systems
Hierarchical Fuzzy Systems (HFSs) have been viewed as a promising option to overcoming a fundamental problem in Fuzzy Logic Systems (FLSs), namely the rule explosion associated with an increase in input variables. In HFSs, the original FLS is decomposed into a number of low-dimensional fuzzy logic s...
Published in: | 2023 IEEE INTERNATIONAL CONFERENCE ON FUZZY SYSTEMS, FUZZ |
---|---|
Main Authors: | , , , , , |
Format: | Proceedings Paper |
Language: | English |
Published: |
IEEE
2023
|
Subjects: | |
Online Access: | https://www-webofscience-com.uitm.idm.oclc.org/wos/woscc/full-record/WOS:001103277400055 |
author |
Razak Tajul Rosli; Kamis Nor Hanimah; Anuar Nurul Hanan; Garibaldi Jonathan M.; Wagner Christian |
---|---|
spellingShingle |
Razak Tajul Rosli; Kamis Nor Hanimah; Anuar Nurul Hanan; Garibaldi Jonathan M.; Wagner Christian Decomposing Conventional Fuzzy Logic Systems to Hierarchical Fuzzy Systems Computer Science |
author_facet |
Razak Tajul Rosli; Kamis Nor Hanimah; Anuar Nurul Hanan; Garibaldi Jonathan M.; Wagner Christian |
author_sort |
Razak |
spelling |
Razak, Tajul Rosli; Kamis, Nor Hanimah; Anuar, Nurul Hanan; Garibaldi, Jonathan M.; Wagner, Christian Decomposing Conventional Fuzzy Logic Systems to Hierarchical Fuzzy Systems 2023 IEEE INTERNATIONAL CONFERENCE ON FUZZY SYSTEMS, FUZZ English Proceedings Paper Hierarchical Fuzzy Systems (HFSs) have been viewed as a promising option to overcoming a fundamental problem in Fuzzy Logic Systems (FLSs), namely the rule explosion associated with an increase in input variables. In HFSs, the original FLS is decomposed into a number of low-dimensional fuzzy logic subsystems. As a result, rules in HFSs typically have antecedents with fewer variables than rules in FLSs which compute similar function mappings, given that the number of input variables of each subsystem is smaller. Consequently, HFSs tend to limit rule explosion, lowering complexity and enhancing model interpretability. However, developing the HFSs is difficult due to the added issue of designing suitable architecture (i.e., various subsystems, levels, topologies, and subsystem interactions) and rules for each subsystem. In fact, decomposing conventional fuzzy system is a challenging task. The difficulties include: How to select the input variable for each subsystem, How to improve the meaning of intermediate variable?, How to link all the subsystems in HFSs?, and How to design the rules for each subsystem? Hence, this paper presents a method to convert conventional FLSs to hierarchical fuzzy systems using two key steps. This method contributes to the process or guidelines in overcoming the difficulties in the decomposition of FLS to HFS. IEEE 1544-5615 2023 10.1109/FUZZ52849.2023.10309727 Computer Science WOS:001103277400055 https://www-webofscience-com.uitm.idm.oclc.org/wos/woscc/full-record/WOS:001103277400055 |
title |
Decomposing Conventional Fuzzy Logic Systems to Hierarchical Fuzzy Systems |
title_short |
Decomposing Conventional Fuzzy Logic Systems to Hierarchical Fuzzy Systems |
title_full |
Decomposing Conventional Fuzzy Logic Systems to Hierarchical Fuzzy Systems |
title_fullStr |
Decomposing Conventional Fuzzy Logic Systems to Hierarchical Fuzzy Systems |
title_full_unstemmed |
Decomposing Conventional Fuzzy Logic Systems to Hierarchical Fuzzy Systems |
title_sort |
Decomposing Conventional Fuzzy Logic Systems to Hierarchical Fuzzy Systems |
container_title |
2023 IEEE INTERNATIONAL CONFERENCE ON FUZZY SYSTEMS, FUZZ |
language |
English |
format |
Proceedings Paper |
description |
Hierarchical Fuzzy Systems (HFSs) have been viewed as a promising option to overcoming a fundamental problem in Fuzzy Logic Systems (FLSs), namely the rule explosion associated with an increase in input variables. In HFSs, the original FLS is decomposed into a number of low-dimensional fuzzy logic subsystems. As a result, rules in HFSs typically have antecedents with fewer variables than rules in FLSs which compute similar function mappings, given that the number of input variables of each subsystem is smaller. Consequently, HFSs tend to limit rule explosion, lowering complexity and enhancing model interpretability. However, developing the HFSs is difficult due to the added issue of designing suitable architecture (i.e., various subsystems, levels, topologies, and subsystem interactions) and rules for each subsystem. In fact, decomposing conventional fuzzy system is a challenging task. The difficulties include: How to select the input variable for each subsystem, How to improve the meaning of intermediate variable?, How to link all the subsystems in HFSs?, and How to design the rules for each subsystem? Hence, this paper presents a method to convert conventional FLSs to hierarchical fuzzy systems using two key steps. This method contributes to the process or guidelines in overcoming the difficulties in the decomposition of FLS to HFS. |
publisher |
IEEE |
issn |
1544-5615 |
publishDate |
2023 |
container_volume |
|
container_issue |
|
doi_str_mv |
10.1109/FUZZ52849.2023.10309727 |
topic |
Computer Science |
topic_facet |
Computer Science |
accesstype |
|
id |
WOS:001103277400055 |
url |
https://www-webofscience-com.uitm.idm.oclc.org/wos/woscc/full-record/WOS:001103277400055 |
record_format |
wos |
collection |
Web of Science (WoS) |
_version_ |
1809678632951480320 |