الفهرس يوجد فقط 14 صفحة متاحة للعرض العام
 |  | from | 84 |  |  |
| | | from | 84 | | |
المستخلص
Demand forecasting is being an urgent field of research nowadays. It is affecting the whole process of supply chain and the performance measures such as inventory levels, customer service satisfaction, and so ort. Most of researchers work(:d on how to use statistical methods in forecasting and how tt> inodify in the methods in order to improve forecasting accura<::y. Combinirtg methbds has been also explored rising moving average or rule based [6] which improved the tbrecasting accuracy but each mddd was wdtking fot st>n1e certain time series.
This thesis proposes a combination model for statistical methods using gdtetic algorithm as searching engine to choose the best combmation of least errdr. Genetic algorithm optimizes also the forecasted activities collections which maximizing the pri::dit. Dividirig the demand factors into controllable and uncontrollable enables and simplifies demand modeling and analysis.
The performarlce of the ptoposed approach is analyzed and compared to the traditit>nal statistical methods tlsed in fPrecasting. The conducted case study showed that the proposed model is imptovlng the fotecast accuracy than the traditiortal methods. The forecasting accuracy and stability measute are improving using the combinational proposed model.
iii