2010 年 9 巻 2 号 p. 431-437
Data mining has been used for obtaining a typical structure of given data set in the field of information science, systems engineering and so on. In this paper, the cocktail-data quantification using the distance mapping learning network architecture is introduced as a kind of data mining for Kansei engineering. And moreover, the network inversion method is newly introduced into the cocktail-data quantification in order to obtain a suitable cocktail based on given user's Kansei information. A simple numerical example is presented with the usage of some typical cocktails and their questionnaire.