On the generalization of redundancy criteria for linguistic fuzzy rules

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F61988987%3A17610%2F20%3AA21025JW" target="_blank" >RIV/61988987:17610/20:A21025JW - isvavai.cz</a>

Výsledek na webu

<a href="https://ieeexplore.ieee.org/document/9287612" target="_blank" >https://ieeexplore.ieee.org/document/9287612</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/KSE50997.2020.9287612" target="_blank" >10.1109/KSE50997.2020.9287612</a>

Jazyk výsledku

angličtina

Název v původním jazyce

On the generalization of redundancy criteria for linguistic fuzzy rules

Popis výsledku v původním jazyce

The aim of this paper is to generalize existing criteria for determining redundant fuzzy linguistic rules, i.e., fuzzy IF-THEN rules that use evaluative linguistic expressions, and that can be freely omitted from a given rule-base without any impact Perception-based logical deduction (PbLD) is a specific inference method tailored to the linguistic fuzzy rules with linguistic expressions and it has been shown that this setting enables the investigation of redundant rules from a theoretical point of view. First of all, rules suspicious from redundancy are identified. Secondly, their elimination has to be checked as the suspicion can be canceled by a presence of a so-called canceling rule. However, the situation is complicated as even the impact of the canceling rule can be eliminated by another rule. The previous studies considered theoretical results and the final algorithm under the assumption of at most a single canceling rule.This paper generalizes the investigation to a more general case when even sets of rules that cancel a single suspicion can be found. We elaborate new criteria for such a purpose that enable to construct an algorithm for an automatic detection of redundant fuzzy linguistic rules. The elimination of redundant rules in a given system is an important step towards simplicity, clarity, and explainability, that moreover, decreases computational costs.

Název v anglickém jazyce

On the generalization of redundancy criteria for linguistic fuzzy rules

Popis výsledku anglicky

The aim of this paper is to generalize existing criteria for determining redundant fuzzy linguistic rules, i.e., fuzzy IF-THEN rules that use evaluative linguistic expressions, and that can be freely omitted from a given rule-base without any impact Perception-based logical deduction (PbLD) is a specific inference method tailored to the linguistic fuzzy rules with linguistic expressions and it has been shown that this setting enables the investigation of redundant rules from a theoretical point of view. First of all, rules suspicious from redundancy are identified. Secondly, their elimination has to be checked as the suspicion can be canceled by a presence of a so-called canceling rule. However, the situation is complicated as even the impact of the canceling rule can be eliminated by another rule. The previous studies considered theoretical results and the final algorithm under the assumption of at most a single canceling rule.This paper generalizes the investigation to a more general case when even sets of rules that cancel a single suspicion can be found. We elaborate new criteria for such a purpose that enable to construct an algorithm for an automatic detection of redundant fuzzy linguistic rules. The elimination of redundant rules in a given system is an important step towards simplicity, clarity, and explainability, that moreover, decreases computational costs.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10102 - Applied mathematics

Projekt

<a href="/cs/project/GA20-07851S" target="_blank" >GA20-07851S: Fuzzy relační struktury v přibližném usuzování</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2020

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

The 2020 12th International Conference on Knowledge and Systems Engineering (KSE)

ISBN

978-172814510-5

ISSN

2164-2508

e-ISSN

—

Počet stran výsledku

6

Strana od-do

136-141

Název nakladatele

IEEE

Místo vydání

Can Tho City

Místo konání akce

Can Tho

Datum konání akce

1. 1. 2020

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—