Using the Backpropagation Algorithm to Distinguish Arabic Alphabet

In this research, a study of the Arabic alphabet used a multi-layered neural network, which is the backpropagation error. Using the algorithm through the Losing activation function to train the network. The hidden numbers of nodes are 10, the number of cycles is 500, and the error is 0.001, using th...

Celý popis

Uložené v:

Podrobná bibliografia
Vydané v:	Zanco Journal of Humanity Sciences Ročník 28; číslo 1
Hlavný autor:	Samyia Khalid Hasan
Médium:	Journal Article
Jazyk:	Arabic English
Vydavateľské údaje:	Salahaddin University-Erbil 14.02.2024
Predmet:	networks neural artificial, back propagation algorithm, weights, input and output
ISSN:	2218-0222, 2412-396X
On-line prístup:	Získať plný text
Tagy:	Pridať tag Žiadne tagy, Buďte prvý, kto otaguje tento záznam!

Popis
Shrnutí:	In this research, a study of the Arabic alphabet used a multi-layered neural network, which is the backpropagation error. Using the algorithm through the Losing activation function to train the network. The hidden numbers of nodes are 10, the number of cycles is 500, and the error is 0.001, using the Matlab R2013a program. The aim of the study It is the use of the network algorithm to recognize the characters, by training the network to recognize the characters in two cases. The first case is inputting the image of the letter into the grid and the second case is identifying the letter that represents the letter drawn in the image. And it was reached that the algorithm used for the network of nervousness to recognize the Arabic alphabet and then show it correctly.
ISSN:	2218-0222 2412-396X
DOI:	10.21271/zjhs.28.1.6