Using neural networls we have simulated a human subjective choice of taste employing as input chemical data and as output the taster's choice of a panel test about the quality of wines and oils. The nets, after training cycles with a few examples, were able to give responses to the quality of all samples with little percentage judgement difference compared to each taster and related to the average Panel's judgement. The nets worked out sets of weights that give out new information showing and quantifying the relevance of each set of input data for the individual taster's choice.
Simulation of Human Edonic Choiches / Romeo, Francesco; Giaccio, Mario. - 1:(1994), pp. 437-440. (Intervento presentato al convegno World Congress on Neural Networks tenutosi a WCNN, San Diego California USA nel 5-9 Giugno 1994).
Simulation of Human Edonic Choiches
Francesco RomeoPrimo
;Mario Giaccio
1994
Abstract
Using neural networls we have simulated a human subjective choice of taste employing as input chemical data and as output the taster's choice of a panel test about the quality of wines and oils. The nets, after training cycles with a few examples, were able to give responses to the quality of all samples with little percentage judgement difference compared to each taster and related to the average Panel's judgement. The nets worked out sets of weights that give out new information showing and quantifying the relevance of each set of input data for the individual taster's choice.File | Dimensione | Formato | |
---|---|---|---|
neural nets Congress_san_diego_vini.pdf
accesso aperto
Descrizione: Articolo
Tipologia:
Documento in Post-print
Licenza:
Dominio pubblico
Dimensione
4.29 MB
Formato
Adobe PDF
|
4.29 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.