Последнее обновление: 25.10.2006   Статьи / Нейронные сети

Нейронные сети

Feedforward Neural Network for Hanwritten Character Recognition Авторы: Janusz A. Starzyk and Nasser Ansari
Организация: Department of Electrical and Computer Engeneering Ohio University, Athens
Дата: 1992 год
Кол-во страниц: 4
Analysis of feedforward neural networks for handwritten character recognition was performed in order to improve the learning capability and accuracy of classification. Training data is first analyzed of line and information which differentiate them most is selected. Selected information determines characteristics of the neural networks. Two algorithms were studied, and it was demonstrated that accuracy similar to the accuracy of backpropogation can be obtained with much less training time and a simple network architecture.
 Скачать файл (321 Кб)


Поддержите проект материально!

рублей Яндекс.Деньгами
на счёт 41001275511292 (cr-online.ru)

Вы используете мой ресурс? Буду вам очень благодарен, если накоплю хотя бы на оплату хостинга.