Abstract:
Multi-criteria inventory classification (MCIC) is a widely used inventory classification method that groups inventory items with respect to several criteria, in order to facilitate their management. Many researchers have been used several methods to solve MCIC problem. However, some of them are quite complex to understand and are not capable of handling qualitative data which is impractical in today's manufacturing conditions. In addition, one of the most common problem is that, in most of the existing methods, when a new inventory item is stored in a warehouse, the classification process must be repeated. In this paper, a new hybrid model generated by genetic algorithm (GA), fuzzy c-means (FCM) and adaptive neuro-fuzzy inference system (ANFIS) is proposed for inventory classification. To create this model, three steps are followed up which are optimizing FCM algorithm by using GA, clustering the data set with FCM algorithm and generating the ANFIS classification model. This model does not need to be regenerated to solve the classification problem whenever a new inventory item is introduced. The model is also capable of handling both quantitative and qualitative criteria. The proposed model is applied to a real-life problem. Results of the model are compared with those of artificial neural network (ANN) model. The comparison showed that the proposed model is more successful than the ANN model.