Title: Application of adaptive neuro-fuzzy inference system and artificial neural network in inventory level forecasting
Authors: Sanjoy Kumar Paul; Abdullahil Azeem; Abhishek Kumar Ghosh
Addresses: Bangladesh University of Engineering and Technology, Dhaka – 1000, Bangladesh ' Bangladesh University of Engineering and Technology, Dhaka – 1000, Bangladesh ' Bangladesh University of Engineering and Technology, Dhaka – 1000, Bangladesh
Abstract: Determining optimum level of inventory is very important for any organisation which depends on various factors. In this research, six main factors have been considered as input parameters and the inventory level has been considered as the single output for this inventory management problem. Price of raw material, demand of raw material, holding cost, setup cost, supplier's reliability and lead time are considered as input parameters. An adaptive neuro-fuzzy inference system (ANFIS) has been applied as the artificial intelligence technique for modelling the inventory problem. ANFIS results have been compared with results from another artificial intelligence technique, artificial neural network (ANN), to validate the output results. Performance of both methods has been shown regarding different error measures. Comparison clearly shows the superiority of ANFIS results over ANN results and thus makes ANFIS a better choice for inventory level forecasting.
Keywords: inventory level forecasting; adaptive neuro-fuzzy inference systems; ANFIS; artificial neural networks; ANNs; fuzzy logic; inventory levels; inventory modelling; raw material prices; raw material demand; holding cost; setup cost; supplier reliability; lead times.
International Journal of Business Information Systems, 2015 Vol.18 No.3, pp.268 - 284
Published online: 28 Mar 2015 *Full-text access for editors Access for subscribers Purchase this article Comment on this article