Conjugate gradient back-propagation with modified Polack–Rebier updates for training feed forward neural network
Joint Authors
al-Bayati, Abbas
Abbu, Khalil K.
Salih, Ibrahim A.
Source
Iraqi Journal of Statistical Science
Issue
Vol. 2011, Issue 20 (31 Aug. 2011), pp.164-173, 10 p.
Publisher
University of Mosul College of Computer Science and Mathematics
Publication Date
2011-08-31
Country of Publication
Iraq
No. of Pages
10
Main Subjects
Topics
Abstract AR
طورت عدة خوارزميات التعليم للشبكات العصبية ذوات التغذية الأمامية و تستند كثير من هذه الخوارزميات الانحدار السلبي، و من المعروف في نظرية الأمثلية أنها ليست ذات كفاءة في التطبيقات العملية.
في هذا البحث نحاول تطويره طريقة بولاك-ريبيير للمتجهات المترافقة لتعليم الشبكة العصبية ذات التغذية الأمامية، و تطويرنا استند إلى معادلة القاطع (شرط شبيه نيوتن).
الخوارزمية المقترحة اختبرت لبعض مسائل الاختبار المعروفة و قورنت مع بعض الخوارزميات المعروفة في هذا المجال.
Abstract EN
Several learning algorithms for feed-forward (FFN) neural networks have been developed, many of these algorithms are based on the gradient descent algorithm well-known in optimization theory which have poor performance in practical applications.
In this paper we modify the Polak-Ribier conjugate gradient method to train feed forward neural network.
Our modification is based on the secant equation (Quasi-Newton condition).
The suggested algorithm is tested on some well known test problems and compared with other algorithms in this field.
American Psychological Association (APA)
al-Bayati, Abbas& Abbu, Khalil K.& Salih, Ibrahim A.. 2011. Conjugate gradient back-propagation with modified Polack–Rebier updates for training feed forward neural network. Iraqi Journal of Statistical Science،Vol. 2011, no. 20, pp.164-173.
https://search.emarefa.net/detail/BIM-399788
Modern Language Association (MLA)
al-Bayati, Abbas…[et al.]. Conjugate gradient back-propagation with modified Polack–Rebier updates for training feed forward neural network. Iraqi Journal of Statistical Science No. 20 (2011), pp.164-173.
https://search.emarefa.net/detail/BIM-399788
American Medical Association (AMA)
al-Bayati, Abbas& Abbu, Khalil K.& Salih, Ibrahim A.. Conjugate gradient back-propagation with modified Polack–Rebier updates for training feed forward neural network. Iraqi Journal of Statistical Science. 2011. Vol. 2011, no. 20, pp.164-173.
https://search.emarefa.net/detail/BIM-399788
Data Type
Journal Articles
Language
English
Notes
Includes bibliographical references : p. 172-173
Record ID
BIM-399788