Modeling of fuzziness in multivariate analysis

This paper introduces a new modeling technique of fuzziness in multivariate analysis based on subjective evaluation data. The subjective evaluation data includes diverse perceptions in the evaluators\´ assessment of the evaluation object. Average data of the evaluators is often used in order to send a generally accurate interpretation of the subjective data. Analyzing of the subjective evaluation data is often called the "Kansei" data analysis. In "Kansei" data analysis, the behavior of the evaluators is often focused on, because of exploring items or objects carrying weight. However, the obtained result by use of the average data does not provide such a character. The proposed modeling technique can provide the tendency of people\´s opinions and at the same time their diversity. This is achieved by extending the traditional multivariate statistical analysis, by use of the fuzzy-sets-theory