| LEADER |
03015nam a22002297a 4500 |
| 001 |
1787074 |
| 024 |
|
|
|3 10.21608/jsfc.2018.85801
|
| 041 |
|
|
|a eng
|
| 044 |
|
|
|b مصر
|
| 100 |
|
|
|9 568424
|a Agamy, Mohamed Abdel Salam
|e Author
|
| 245 |
|
|
|a A Proposed Method for the Analysis of Multiple Regression Using Artificial Intelligence
|
| 260 |
|
|
|b جامعة الأزهر - كلية التجارة
|c 2018
|g يناير
|
| 300 |
|
|
|a 75 - 113
|
| 336 |
|
|
|a بحوث ومقالات
|b Article
|
| 520 |
|
|
|b This research presents a proposed method to overcome difficulties which we face when we use traditional regression analysis method and provide solutions in an automated manner that does not require experience or a large field study. Some of these difficulties are mathematical difficulties, some of which are the need for considerable experience to choose the appropriate data model, especially when the traditional regression analysis method is not available. The new approach raises forecast accuracy with powerful Neural Network MATLAB code. This approach automatically trains neural networks and applies them for regression analysis, thus getting accurate, business or stock market predictions doesn’t require much effort or time. This means faster and more precise results than ever before. This research aims to: 1. Introducing a new method for the analysis of multiple regression using artificial neural networks, which represent one of the most important areas of artificial intelligence. 2. Comparing the suggested method of artificial neural networks with the traditional method of multiple regression analysis to determine which is better for prediction. 3- Conducting an applied study using realistic data. 4 - Formulation of a computer program using MATLAB software packages to compare the two methods. 5. Formulate a program using MATLAB software packages to predict regression models using both neural networks and conventional methods It has become clear through the applied study that: a- Proposed method can be used to model data and predict its unknown values. b- The proposed method using neural networks is better than traditional in the case of nonlinear models. c- Performance of the proposed method is better than traditional in the case of small samples. d- Convergence of the results of the two methods of increasing sample sizes.
|
| 653 |
|
|
|a علم الإحصاء
|a التحصلص الإحصائي
|a الذكاء الاصطناعي
|a الشبكات العصبية
|
| 700 |
|
|
|a Wahba, Mahmoud El-Sayed
|e Co-Author
|9 568425
|
| 773 |
|
|
|4 الاقتصاد
|4 إدارة الأعمال
|6 Economics
|6 Management
|c 013
|e Scientific Journal of the Faculty of Commerce
|f Al-Maǧallah al-ʿilmiyyaẗ li Qitāʿ Kulliyyāẗ al-tiǧāraẗ bi Ǧāmiʿaẗ al-Azhar
|l 019
|m ع19
|o 0797
|s المجلة العلمية لقطاع كليات التجارة
|v 000
|x 2636-3674
|
| 856 |
|
|
|n https://jsfc.journals.ekb.eg/article_85801.html
|u 0797-000-019-013.pdf
|
| 930 |
|
|
|d y
|p y
|q n
|
| 995 |
|
|
|a EcoLink
|
| 999 |
|
|
|c 1049542
|d 1049542
|
