Evaluation of Customer-Oriented Service Quality on Self-Adaptive RBFNN and UDM

Evaluation of customer-oriented service quality (ECSQ) is an important practical measure to increase enterprises' competition capacity. In this article, a performance evaluation indexes system is established. UDM and fussy-mathematics are used to generate enough samples to train self-adaptive RBFNN. The instances of applying RBFNN to do ECSQ shows that its generalization capacity is more powerful than general RBFNN. Meanwhile, the randomicity of former fussy-comprehensive evaluation and man-made uncertainty are conquered.