Decision Support System In The Vending Machine Management

Nowadays, the vending manchine market in China is in its fast growing stage. The foundation technologies of vending machines have been mature as monitoring system, sales system, payment system and so on. The operation and management of the vending machine will be the focus of each manufacturer competition. Therefore, designing a decision support system for the managers in making management decision and other assisted support through a number of intelligent algorithms is necessary.Focus on the prediction and decision-making of the decision support system of the vending machine based on the understanding of the decision support systems and vending machines work, and the corresponding application development in this paper. The main work is as follows:(1) Describes the development and application of the decision support system and briefs some technolegy will be used in this paper.(2) Proposes a prediction scheme for the sales of commodities on the vending machine, predicts the sales based on time series analysis, neural networks and Holt-Winters theory using historical sales data. Comparing the performance of each method to obtain the reliable predictions.(3) As historical sales data is subject to the bad sales program, we propose a cross prediction method to make up the missing data, which making the prediction more accurace.(4) Imports the decision tree algorithm and takes sales, profits, the cost of space, alternative commodities such factors as the dicriminating attributes to achieve the development of sales program by classifying and rating the commodity.(5) Builds decision support system include such module functions like human-computer interaction, data management, model management.etc by Java, and the open source software R completes the calculations and statistics. Finally acquires the accurate predict results and classfications of commodities by competing the connection between Java and R, which provide an important imformantion to the managers.