Democritus University of Thrace (DUTH)
Μαθηματική μοντελοποίηση με χρήση της ασαφούς λογικής με εφαρμογές σε θέματα μηχανικών
Abstract
dc:descriptionFuzzy logic is the generalization of the classical two-valued logic and as an effect the classical set theory. The ability of fuzzy logic to describe sets and systems, in which ambiguity is present, makes fuzzy logic a fundamental part of computational intelligence. In particular, fuzzy logic and fuzzy set theory techniques, such as fuzzy systems, are implemented in the fields of artificial intelligence, approximate reasoning and inference systems. Fuzzy systems are employed in fuzzy logic engineering applications and they are based on fuzzy set construction, the use of fuzzy operators and the implementation of fuzzy inference rules. These fuzzy rules are described by functions, which are relations between fuzzy sets that stem from fuzzy propositional theory. More specifically, such functions, which are essential for the inference making process involved in fuzzy systems mechanisms, are usually fuzzy implications. In data applications in which fuzzy implications are involved, the selection of the most suitable fuzzy implication or a set of suitable fuzzy implications is of major importance, since the induction of the system in which are included is affected by this selection. The problem of choosing suitable fuzzy implications for data applications does not have a general solution. Moreover, the theoretical suitability is not derived from the dataset of an application but from the criteria that ensue from the theoretic rules of inference. These problems became the motive to produce the method this research refers to. The scope of this dissertation is the development of a method for the detection of the most suitable fuzzy implication given an initial set of fuzzy implications that are to be engaged in fuzzy systems. The result of the method is the fuzzy implication which is the most suitable for implementing the fuzzy inference rules of a fuzzy system regarding the dataset of an application. In parallel, a new approach of evaluating the fuzzy implication suitability is proposed with which the detection of the most suitable one is achieved concerning the data of the application. Additionally, the fuzzy conditional propositions involving the fuzzy implication, which is the output of this method, are the ones with the highest truth values for the measurements of the application. As a result, the method detects the fuzzy implication which best reflects the notion of induction of the dataset of the application, in comparison to the rest of the fuzzy implications under consideration.
Degree
thesis:*- Grantor dc:publisher
- Democritus University of Thrace (DUTH)
- Year dc:date
- 2019
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Pagouropoulos, Panagiotis
- Παγουρόπουλος, Παναγιώτης
Subjects
dc:subject × 20- Ασαφείς συνεπαγωγές
- Ασαφείς προτάσεις
- Λεκτικές μεταβλητές
- Μέτρα ομοιότητας
- Απόσταση ασαφών συνόλων
- Fuzzy implications
- Fuzzy propositions
- Linguistic variables
- Similarity measures
- Fuzzy set distance
- Φυσικές Επιστήμες
- Μαθηματικά
- Επιστήμες Μηχανικού και Τεχνολογία
- Περιβαλλοντική Μηχανική
- Άλλες Επιστήμες Μηχανικού και Τεχνολογίες
- Natural Sciences
- Mathematics
- Engineering and Technology
- Environmental Engineering
- Other Engineering and Technologies
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/46280