Automated system of recognizing the printable characters by neural networks of Hopfield and Hamming
Автор: Lyozina Irina, Mazayev Alexey
Журнал: Известия Самарского научного центра Российской академии наук @izvestiya-ssc
Рубрика: Перспективные информационные технологии
Статья в выпуске: 2-5 т.17, 2015 года.
Бесплатный доступ
The article describes an automated system developed for recognizing the printed symbols. The system uses recurrent neural networks of Hamming and Hopfield. The first one is trained by the standard learning algorithm, the second one - by the Hebb's rule and delta-projections. The article analyzes the results of the recognizing.
Метод delta-проекций, automated system, neural network, hamming network, hopfield network, pattern recognition, printable character, hebb's rule, the method of delta-projections
Короткий адрес: https://sciup.org/148203700
IDR: 148203700
