Abstract Fuzzy regression models often consider triangular or trapezoidal membership functions. Of course this kind of fuzzy numbers has the advantage that they are much easy to use; nevertheless several doubts arise about the appropriateness of the triangular shape. As known, the fuzzy sets are very useful when there are imprecise data and in particular when, within an interval, the degree of truth of the values is not always the same. Fuzzy numbers are also useful to translate terms of human language into numbers. Triangular fuzzy numbers offer a good solution for dealing with this kind of data. However in this paper it is proposed an alternative membership function which appears to be much more appropriate; it has been called ”fuzzy number associated with the normal curve”. In this framework it will be justified the introduction of this membership function as a valid substitute for triangular fuzzy numbers.
Bell Shaped Fuzzy Numbers Associated With The Normal Curve
MATURO, FABRIZIO
;FORTUNA, FRANCESCA
2014-01-01
Abstract
Abstract Fuzzy regression models often consider triangular or trapezoidal membership functions. Of course this kind of fuzzy numbers has the advantage that they are much easy to use; nevertheless several doubts arise about the appropriateness of the triangular shape. As known, the fuzzy sets are very useful when there are imprecise data and in particular when, within an interval, the degree of truth of the values is not always the same. Fuzzy numbers are also useful to translate terms of human language into numbers. Triangular fuzzy numbers offer a good solution for dealing with this kind of data. However in this paper it is proposed an alternative membership function which appears to be much more appropriate; it has been called ”fuzzy number associated with the normal curve”. In this framework it will be justified the introduction of this membership function as a valid substitute for triangular fuzzy numbers.File | Dimensione | Formato | |
---|---|---|---|
bell.pdf
accesso aperto
Descrizione: Procedings
Tipologia:
Documento in Post-print
Dimensione
4.64 MB
Formato
Adobe PDF
|
4.64 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.