Integration of Data Mining and Stochastic Dynamic Programming to Present a Research Framework for After-sales Service

Manufacturing firms can obtain considerable income and profit via After Sales Services. This paper presents a framework consists of Data Mining technology and Dynamic Programming to analyze the different relationships among automotive part failures and then to determine the optimal decisions in case of part replacement. The proposed model at first collect the proper data records and after preprocessing, in two separate streams looks for significant relationships and association rules. K-Means and Generalized Rule Induction algorithms are used to determine the relationship between part failures and different variables. Then Market Basket Analysis is applied to find a set of dependent components. The process of replacing these components is modeled as a Markov process and is solved by Stochastic Dynamic Programming. The effectiveness of the model is illustrated with the warranty data mining application from the automotive industry in Iran.