Abstract
This paper proposes a method for obtaining a final list of raw materials requirements, which must be available in the warehouse of automotive companies that produce rubber and rubber-metal parts for manufacturing engines, gearboxes, shock absorbers, etc. This method determines the critical materials considering the historical records of sales. For each critical item, a model of distribution of sales is determined using classical probabilistic models and Finite Mixtures of Densities, decision criteria for choosing the best model was based on the Akaike Information Criterion value. Also, for each critical item a sales forecast is performed using five of the most common methods, choosing the best forecast with the lowest mean absolute deviation (MAD) value. Finally, the service level to supply raw materials for the company is determined. All the calculations, estimation of parameters of each model, forecasts, etc. were made using the R statistical software.
Original language | English |
---|---|
Pages | 2049-2058 |
Number of pages | 10 |
State | Published - 2013 |
Event | IIE Annual Conference and Expo 2013 - San Juan, Puerto Rico Duration: 18 May 2013 → 22 May 2013 |
Conference
Conference | IIE Annual Conference and Expo 2013 |
---|---|
Country/Territory | Puerto Rico |
City | San Juan |
Period | 18/05/13 → 22/05/13 |
Keywords
- AIC value
- Finite mixture models
- Gamma distribution
- Supply