Thesis MODELOS DE ESTIMACIÓN DE LA DEMANDA PARA EMBONOR COCA-COLA S.A
Date
2017
Authors
KARAKI AHMAD, SALAH
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
En los últimos años Embonor S.A ha aumentado el nivel de demanda insatisfecha. Una de las causas atribuibles es que la programación de la producción no se está llevando a cabo óptimamente. La programación de la producción en cada línea productiva se realiza mediante el criterio de cobertura de stock de cada producto. El ratio de cobertura es determinado a través de una estimación de la demanda. Actualmente, se estima la demanda para todos los productos (84 ‘SKU’s’) con el mismo modelo de predicción. En el presente estudio se evalúa 53 modelos de predicción de demanda para cada ‘SKU’. Para cada producto se determina el modelo que minimiza la cuasivarianza muestral de los errores de pronóstico, determinándose de esta forma el mejor método de predicción para cada producto.En este estudio se determinó que el modelo de predicción actual de la empresa es válido solo en seis de los 84 ‘SKU’s’, i.e., en este estudio se identificaron mejores métodos de pronóstico para 78 ‘SKU’s’.
In recent years Embonor S.A has increased the level of unsatisfied demand. One of the attributable causes is that the scheduling of production is not being carried out optimally. The scheduling of production in each production line is done using the criterion of stock coverage of each product. The coverage ratio is determined through an estimate of demand. Currently, the demand for all products (84 'SKU's') is estimated using the same prediction model. The present study evaluates 53 demand prediction models for each 'SKU'. For each product, the model that minimizes the quasivariance of the forecast errors is determined, thus determining the best prediction method for each product.In this study it was determined that the company's current prediction model is valid only in six of the 84 'SKU's', i.e., in this study better forecasting methods were identified for 78 'SKU's'.
In recent years Embonor S.A has increased the level of unsatisfied demand. One of the attributable causes is that the scheduling of production is not being carried out optimally. The scheduling of production in each production line is done using the criterion of stock coverage of each product. The coverage ratio is determined through an estimate of demand. Currently, the demand for all products (84 'SKU's') is estimated using the same prediction model. The present study evaluates 53 demand prediction models for each 'SKU'. For each product, the model that minimizes the quasivariance of the forecast errors is determined, thus determining the best prediction method for each product.In this study it was determined that the company's current prediction model is valid only in six of the 84 'SKU's', i.e., in this study better forecasting methods were identified for 78 'SKU's'.
Description
Catalogado desde la version PDF de la tesis.
Keywords
EMBONOR COCA-COLA S.A , ESTIMACION DE DEMANDA , MODELO DE PREDICCION