Article Per Year (5 Year)
p-Index From 2021 - 2026
0
P-Index
This Author published in this journals
All Journal Media Aplikom
Nahar Mardiyantoro
Sekolah Tinggi Ilmu Komputer Yos Sudarso Purwokerto
Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Library & Information Science

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Metode Ekstraksi Data Untuk Pengenalan Huruf Dan Angka Tulisan Tangan Dengan Menggunakan Jaringan Syaraf Buatan Propagasi Balik Nahar Mardiyantoro; Muhammad Hery Santoso
Media Aplikom Vol 1 No 3 (2010)
Publisher : Sekolah Tinggi Ilmu Komputer YOS SUDARSO Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

In this paper presented a method to identify patterns of handwritten letters and numbers using the concept of back propagation artificial neural network. Used artificial neural network using sigmoid activation function at all la-pisannya, as well as neuron structure is rigid (not flexible). The pattern recognition method of handwritten letters and numbers tersebutdifokuskan on how the data extraction from existing samples. The steps adalahpertama data extraction, dividing the sample into several areas of observation (region). Second, from each region were taken pixelyang active value, so that numerical data obtained by the existing jumlahregion. Third, the numerical data that have been generated, dinormali-sasi fixed by comparison, where each numerical data from each region was divided by the largest value of all numerical data from the same sample. To prove the method which the author makes a aplikasipendukung Char-Cog-nitrones by using Delphi 6.0 Enterprise. As for the steps-analisaan lawyer who first do is determine the best artificial neural network characteristics for this method. Then, analyzed the composition of the best regions for the sample to be used so as to provide the best end result. The final results of the analysis showed that the patterns produced by using these methods can be recognized well by the artificial neural network.
Co-Authors Muhammad Hery Santoso