Abstract
Scientific cost prediction is the foundation for formulating a business plan, which helps enterprises to determine the target costs and to control the operating costs. In today's world, quantitative methods are crucial for prediction purposes in cost management. On the basis of studying some typical methods of cost prediction, we propose a cost prediction method based on an improved fuzzy model. Firstly, owning to the inefficiency of existing fuzzifier, such that it is difficult to determine the number of fuzzy rules, we present an improved fuzzy c-means (FCM) fuzzifier to solve this problem referring to "granulation-degranulation" strategy and sample error function. Secondly, a fuzzy model based on the improved FCM fuzzifier is constructed through combining Takagi-Sugen-Kang fuzzy inference system. The algorithm of the improved fuzzy model is also realized by this study. Finally, through a series of experimental analyses for steel number product and steel class product of certain machinery enterprise, the results have demonstrated that the proposed method has the better prediction accuracy, which can be used effectively and practically in real-world cost prediction.