Data mining based on local Association Network: The simulation and test on inventory data

This paper proposes a new method of data mining called Association Network (AE), combining the association analysis and sequential pattern. After having AE positioned in terms of time and quantity, it could be used to control inventory without main plans. AE is proved to be able to improve the efficiency of inventory control though simulation and test on inventory data. Besides, this paper also builds a framework of decision support system which can improve the performance of inventory control by making full use of AE.