A fuzzy ordinary regression method for modeling customer preference in tea maker design

Abstract

; The telephone number is (618) 92662945 and the fax number is 618 92662819.2 C.K. Kwong's email address is c.k.kwong@polyu.edu.hk. The telephone number is (852) 27666610 and the fax number is (852) 23625267.3 M.C. Law's email address is daniel@gewcorp.com. The telephone number is (852) 2343 8211 and the fax number is (852) 23431647. Abstract-Faced with fierce competition in marketplaces, manufacturers need to determine the appropriate settings of engineering characteristics of the new products so that the best customer preferences of the products can be obtained. To achieve this, functional models relating customer preferences to engineering characteristics need to be developed. As information regarding functional relationships between customer preferences are generally subjective or heuristic in nature, development of the customer preference models involve two uncertainties, namely fuzziness and randomness. Existing approaches use only fuzzy-based technologies to address the uncertainty caused by fuzziness. They are not designed to address the randomness of the observed data which is caused by a limited knowledge of the variability of influences between customer preferences and engineering characteristics. In this article, a fuzzy ordinary regression method is proposed to develop the customer preference models which are capable of addressing the two uncertainties of crispness and fuzziness of the customer preferences. A case study of a tea maker design which involves both uncertainties is used to demonstrate the effectiveness of the proposed method.

A forward selection based fuzzy regression for new product development that correlates engineering characteristics with consumer preferences

“…P-FR has been used to develop consumer preference models for mobile phone design [22]. c) Hybrid fuzzy least square regression (H-FLSR) [2] can be used to address the uncertainties caused by fuzzy and random natures of the samples. H-FLSR was used to develop consumer preference models for packing machines [17].…”

“…Also, the consumer preference model attempts to address uncertainties caused by the quantitative evaluations and measures of consumer preferences. As (2) is involved with linear terms, linear correlation between consumer preference and engineering characteristics can be addressed by the linear terms, 1 z , 2 z ,…, and m z .…”

“…In order to study the fuzziness of evaluating the three consumer preferences, the experiments configured with L 16 (4 5 ) were repeated twice; the fuzzy observations for catechin content, tea concentration, and tea temperature are shown in Table 3. The details of the experiments can be referred to [2].…”

“…When the correlation between consumer preferences and engineering characteristics is developed, a successful new product is more likely to be produced as consumer requirements can be satisfied based on the truly specified engineering characteristics [2]. Based on quality function deployment (QFD) [1,3], these correlations can be represented by a matrix, namely houses of quality (HOQ), which relates consumer preferences to engineering characteristics.…”

A Stepwise-Based Fuzzy Regression Procedure for Developing Customer Preference Models in New Product Development

“…Therefore, the 5 engineering characteristics which are significant for customer preferences are given as: reheating temperature (x 1 ), number of drops for the first brewing (x 2 ), dipping time (x 3 ), number of drops for the second brewing (x 4 ), and immersion time in the second brewing (x 5 ). The above five steps are detailed in [57].…”