A neuro-fuzzy approach to generating customer satisfaction model for new product development

Understanding customer perception towards consumer products is of extremely important to design teams for designing new products. It is because success of new products is heavily dependent on the associated customer satisfaction level. If the consumers are satisfied with a new product, the chance of the product to be successful in a marketplace would be higher. In this study, we applied adaptive neuro-fuzzy inference system (ANFIS) to generate customer satisfaction models based on market survey data. A modified ANFIS (M-ANFIS) is proposed by which explicit customer satisfaction models can be generated. The models can efficiently deal with continuous input values instead of crispy numbers. To justify M-ANFIS, it was compared with a well-known statistical method, multiple linear regression (MLR). Experimental results indicated that the M-ANFIS outperformed MLR in terms of mean absolute errors and variance of errors.