Daily Demand Forecast Using Artificial Intelligence Techniques for Refectory

Nowadays many organization outsource food services from professional catering company. They pay the bill of catering services as the amount their food demand. Despite changing daily consumption amount according to types of the food, day, behaviors of company employees and number of employees, a decision which is about amount of food demand is made by personal judgement in most organization and thus serviced foods are run to waste. In this study, the data of a university’s refectory is obtained and artificial neural networks, support vector machine and linear regression methods in WEKA tool are used to predict amount of daily food consumption. Also performance of the prediction methods are presented in this study. This paper proposes that different methods are dominant according to the type of consumer with different behaviors such as the university has student, lecturer and other staff and selected method by artificial intelligence techniques provides that the amount of waste food can be minimized