|
Document Type |
: |
Thesis |
|
Document Title |
: |
Forecasting sales revenue by artificial neural network methods in the airline industry: A case study in Saudi Arabia توقع عائدات المبيعات بإستخدام طرق الشبكات العصبية الاصطناعية في صناعة الطيران: دراسة حالة بالمملكة العربية السعودية |
|
Subject |
: |
Faculty of Engineering |
|
Document Language |
: |
Arabic |
|
Abstract |
: |
Forecasting plays an important role in the operations of modern management. It is an important and necessary aid to planning which is the backbone of effective operations. Many organizations have lost revenue, demand, or market share because of poor forecasting. There are many ways to estimate the future. In numerous firms, the entire process is subjective, involving intuition or developed based on staff’s experience. However there are also many quantitative forecasting models such as moving average, exponential smoothing, trend projections, and neural networks. This thesis aims to utilize artificial neural network methods in forecasting sales revenue of an airline company in KSA. The study uses a set of historical data from two large sectors that the airline is operating in. These sectors are Jeddah-Cairo market and Jeddah-Istanbul markets. In the study, initially data are collected, and the ANN models were built using feedforward network architecture and backpropagation as a learning algorithm. After the completing training and learning phases, ANN models outcomes were obtained that can deliver the desired results. The study finalized ANN models that can be used effectively for forecasting sales revenue. The study also introduces a comparison between the forecasting results of ANN and the forecasting results of scientific techniques. In both sectors under study; ANN provided accuracy better than other estimate techniques such as moving average and exponential smoothing. It is found from this research that if markets are having some similarities, the ANN models can be generalized and used for different markets. Also, the study found that whether the parameters are selected based on correlation tests or experience the forecasting results will be almost the same. This study concluded that ANN have been shown to be very promising for forecasting the airlines sales revenue due to their ability to learn from past data, and the ability to generalize. The approach of this study also concluded that the most basic neural network models can outperform the traditional forecasting methods such as moving average and exponential smoothing. |
|
Supervisor |
: |
Prof. Osman E. Taylan |
|
Thesis Type |
: |
Master Thesis |
|
Publishing Year |
: |
1439 AH
2018 AD |
|
Added Date |
: |
Monday, May 28, 2018 |
|
Researchers
| رائد أحمد كلنتن | Kalantan, Raed Ahmad | Researcher | Master | |
|