Title: Demand forecasting model for slow moving items
Authors: Ekambaram Dhakshayani; S. Narayanan
Addresses: School of Mechanical and Building Sciences (SMBS), VIT University, Vellore, Vellore Dt-632014, Tamilnadu, India ' VIT University, Vellore, Vellore Dt-632014, Tamilnadu, India
Abstract: Demand forecasting for a slow moving product is a critical area of concern and becomes crucial to find the optimistic forecast value for a slow moving product which results in huge inventory ultimately affecting the cost. The two new models namely probabilistic demand model (PDM) and demand size-based model (DSBM) have been developed in this work by modifying the Croston's model. Probabilistic demand model (PDM) has been developed by considering actual probabilities in the demand size per period and demand size-based model (DSBM) by directly estimating the demand size without any average per period considerations. Comparative study with the commonly used demand forecasting methods shows that the models follow close approximations to them and are versatile with different values of smoothing constant.
Keywords: probabilistic demand models; modelling; demand size-based models; smoothing constant; Croston; demand forecasting; demand history; irregular demand; intermittent demand; mean error; mean absolute deviation; slow moving items; probability.
DOI: 10.1504/IJLSM.2014.064660
International Journal of Logistics Systems and Management, 2014 Vol.19 No.2, pp.245 - 260
Received: 06 Jul 2013
Accepted: 23 Jul 2013
Published online: 13 Sep 2014 *