Quality modeling for coffee using the knowledge information processing

Taizo Hanai, Eiji Ohkusu, Hiroyuki Honda, Fumio Ito, Motohiko Sugiura, Ichiro Asano, Takeshi Kobayashi

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

MRA (multiple regression analysis), NN (neural network) and FNN (fuzzy neural network) were applied in order to construct the models estimated from the analysis data for the sensory evaluations of various coffee samples. The estimated values using the MRA model for several sensory evaluations did not coincide well with the real values. NN and FNN models had high precision for sensory evaluations except for the sensory evaluation of "Hard". Membership functions and production rules, which were understandable qualitatively, were obtained from the constructed FNN model.

Original languageEnglish
Pages (from-to)560-568
Number of pages9
JournalNippon Shokuhin Kagaku Kogaku Kaishi
Volume44
Issue number8
DOIs
Publication statusPublished - Jan 1 1997
Externally publishedYes

Fingerprint

Coffee
Automatic Data Processing
neural networks
sensory evaluation
Neural Networks (Computer)
Regression Analysis
regression analysis
data analysis
sampling

All Science Journal Classification (ASJC) codes

  • Food Science

Cite this

Hanai, T., Ohkusu, E., Honda, H., Ito, F., Sugiura, M., Asano, I., & Kobayashi, T. (1997). Quality modeling for coffee using the knowledge information processing. Nippon Shokuhin Kagaku Kogaku Kaishi, 44(8), 560-568. https://doi.org/10.3136/nskkk.44.560

Quality modeling for coffee using the knowledge information processing. / Hanai, Taizo; Ohkusu, Eiji; Honda, Hiroyuki; Ito, Fumio; Sugiura, Motohiko; Asano, Ichiro; Kobayashi, Takeshi.

In: Nippon Shokuhin Kagaku Kogaku Kaishi, Vol. 44, No. 8, 01.01.1997, p. 560-568.

Research output: Contribution to journalArticle

Hanai, T, Ohkusu, E, Honda, H, Ito, F, Sugiura, M, Asano, I & Kobayashi, T 1997, 'Quality modeling for coffee using the knowledge information processing', Nippon Shokuhin Kagaku Kogaku Kaishi, vol. 44, no. 8, pp. 560-568. https://doi.org/10.3136/nskkk.44.560
Hanai, Taizo ; Ohkusu, Eiji ; Honda, Hiroyuki ; Ito, Fumio ; Sugiura, Motohiko ; Asano, Ichiro ; Kobayashi, Takeshi. / Quality modeling for coffee using the knowledge information processing. In: Nippon Shokuhin Kagaku Kogaku Kaishi. 1997 ; Vol. 44, No. 8. pp. 560-568.
@article{a1a926e4273b439cbc60c6e60d83fa68,
title = "Quality modeling for coffee using the knowledge information processing",
abstract = "MRA (multiple regression analysis), NN (neural network) and FNN (fuzzy neural network) were applied in order to construct the models estimated from the analysis data for the sensory evaluations of various coffee samples. The estimated values using the MRA model for several sensory evaluations did not coincide well with the real values. NN and FNN models had high precision for sensory evaluations except for the sensory evaluation of {"}Hard{"}. Membership functions and production rules, which were understandable qualitatively, were obtained from the constructed FNN model.",
author = "Taizo Hanai and Eiji Ohkusu and Hiroyuki Honda and Fumio Ito and Motohiko Sugiura and Ichiro Asano and Takeshi Kobayashi",
year = "1997",
month = "1",
day = "1",
doi = "10.3136/nskkk.44.560",
language = "English",
volume = "44",
pages = "560--568",
journal = "Nippon Shokuhin Kagaku Kogaku Kaishi",
issn = "1341-027X",
publisher = "Japanese Society for Food Science and Technology",
number = "8",

}

TY - JOUR

T1 - Quality modeling for coffee using the knowledge information processing

AU - Hanai, Taizo

AU - Ohkusu, Eiji

AU - Honda, Hiroyuki

AU - Ito, Fumio

AU - Sugiura, Motohiko

AU - Asano, Ichiro

AU - Kobayashi, Takeshi

PY - 1997/1/1

Y1 - 1997/1/1

N2 - MRA (multiple regression analysis), NN (neural network) and FNN (fuzzy neural network) were applied in order to construct the models estimated from the analysis data for the sensory evaluations of various coffee samples. The estimated values using the MRA model for several sensory evaluations did not coincide well with the real values. NN and FNN models had high precision for sensory evaluations except for the sensory evaluation of "Hard". Membership functions and production rules, which were understandable qualitatively, were obtained from the constructed FNN model.

AB - MRA (multiple regression analysis), NN (neural network) and FNN (fuzzy neural network) were applied in order to construct the models estimated from the analysis data for the sensory evaluations of various coffee samples. The estimated values using the MRA model for several sensory evaluations did not coincide well with the real values. NN and FNN models had high precision for sensory evaluations except for the sensory evaluation of "Hard". Membership functions and production rules, which were understandable qualitatively, were obtained from the constructed FNN model.

UR - http://www.scopus.com/inward/record.url?scp=0001643577&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0001643577&partnerID=8YFLogxK

U2 - 10.3136/nskkk.44.560

DO - 10.3136/nskkk.44.560

M3 - Article

AN - SCOPUS:0001643577

VL - 44

SP - 560

EP - 568

JO - Nippon Shokuhin Kagaku Kogaku Kaishi

JF - Nippon Shokuhin Kagaku Kogaku Kaishi

SN - 1341-027X

IS - 8

ER -