To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
Forecasting Daily Demand of Orders Using Random Forest Classifier |
|
Author |
Ahmed Alsanad |
| Citation |
Vol. 18 No. 4 pp. 79-83 |
|
Abstract |
In logistics companies, forecasting daily demand of orders is crucial in scheduling and planning tasks of supply chain to meet the consumer needs on time, improving the efficiency and reducing the costs. Even though there are some machine learning techniques have been used for predicting daily demand orders of products in logistics companies for supply chain, the choice of the most appropriate forecasting method remains a significant concern. In this paper, we investigate the application of random forest (RF) for predicting the daily demand orders of products in short time interval. We chose RF classifier in our methodology because it is a sophisticated machine learning technique that faces the strain between over-fitting and under-fitting circumstances. The methodology is evaluated on a real database of a Brazilian logistics company collected during 60 days. The RF classifier is trained on this collected dataset using 10-folds cross validation mode to predict the daily demand of orders of 6 days 10 times. The experiment show the ability the proposed classifier to predict the daily demand of orders with a high accuracy result compared to the baseline classifiers in the state-of-the-art. |
|
Keywords |
Forecasting, Daily demand of orders, Logistics companies, Supply chain, Machine learning, Random forest classifier. |
|
URL |
