Demand Forecasting Algorithms and Supply Chain Optimization for Fresh and Perishable Biological Foods

Supply chain optimization is a crucial issue in many real business con-texts. The main criticalities stem from lack of information sharing amongthe actors involved in the management process and from the uncertaintyon future sales. In this study we propose an integrated approach thataccounts for the following issues: demand forecasting, order planning anddelivery optimization. To this aim, we use three alternative forecastingmodels, propose a three-step procedure for model tuning, and an algo-rithmic approach to support the model choice. Sales forecast is treated asproxy of expected demand and used as input for a multi-objective opti-mization model in order to determine the order plans with respect to fourdifferent criteria, i.e., shortage, outdating, freshness and stock of prod-ucts. On the basis of the selected order plan, a Mixed-Integer LinearProgramming model is devoted to solve the delivery planning problem.The proposed method is applied to a set of real data; the results obtainedare reported and discussed through some examples.