Dijous 13 de Novembre, 13:00 hores, Saló de Graus.
In a process of selecting a correct model to predict and plan an optimal policy for an industry, a Bayesian approach is useful, especially when there are not a lot of data. In those cases, the user is faced with possibilities which could lead to imprecise estimations, and it could cause a lack of stock for distribution, or the opposite, an excess of inventories, both carrying loses for the company. We show a predictive model with bayesian inference which help us to create an optimal policy to plan inventories for a confection industry.