Development of quality prediction models for roasted and ground coffee products | การพัฒนาแบบจำลองการทำนายคุณภาพของผลิตภัณฑ์กาแฟคั่วบด
2009
Siriyaporn Kraimark(Kasetsart University, Bangkok (Thailand). Faculty of Agro-Industry. Department of Agro-Industry Technology Management) E-mail:[email protected] | Parthana Parthanadee(Kasetsart University, Bangkok (Thailand). Faculty of Agro-Industry. Department of Agro-Industry Technology Management) E-mail:[email protected] | Jirachai Buddhakulsomsiri(Thammasat University. Rangsit Center, Pathum Thani (Thailand). Sirindhon International Institute of Technology) E-mail:[email protected]
This paper presented a development of prediction models for the quality evaluation of a roasted and ground coffee products by using regression analysis and artificial neural network techniques. The models were constructed using twenty-one independent variables which were storage duration, color, oiliness, aroma, and size of coffee beans, seven sources of Robusta coffee beans and nine sources of Arabica coffee beans. The dependent variables were the sensory evaluation scores of color, flavor, bitterness, taste and taste balance of the coffee products. From the experiments, the best prediction models were the artificial neural network model with a momentum of 0.4, a learning rate of 0.25, 9 hidden neural nodes, a hyperbolic tangent activation function and an initial weight of 5 for the factor group of taste scores and the artificial neural network model with a momentum of 0.4, a learning rate of 0.25, 6 hidden neural nodes, a hyperbolic tangent activation function and an initial weight of 2 for the group of other factors' scores.
Afficher plus [+] Moins [-]Mots clés AGROVOC
Informations bibliographiques
Cette notice bibliographique a été fournie par Kasetsart University
Découvrez la collection de ce fournisseur de données dans AGRIS