APLIKASI JARINGAN SYARAF TIRUAN MENGGUNAKAN METODE LEARNING VECTOR QUANTIZATION DALAM PENGENALAN POLA HURUF PEGON JAWA

HARI SURRISYAD, NIM. 13650014 (2017) APLIKASI JARINGAN SYARAF TIRUAN MENGGUNAKAN METODE LEARNING VECTOR QUANTIZATION DALAM PENGENALAN POLA HURUF PEGON JAWA. Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (APLIKASI JARINGAN SYARAF TIRUAN MENGGUNAKAN METODE LEARNING VECTOR QUANTIZATION DALAM PENGENALAN POLA HURUF PEGON JAWA) 13650014_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf Download (3MB) | Preview
	Text (APLIKASI JARINGAN SYARAF TIRUAN MENGGUNAKAN METODE LEARNING VECTOR QUANTIZATION DALAM PENGENALAN POLA HURUF PEGON JAWA) 13650014_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf Restricted to Registered users only Download (7MB)

Abstract

Technological development Provides many options how to process data into useful information, One branch of the science of artificial intelligence is Artificial Neural Network which is one of the information processing system designed similar to the performance of the human brain. Artificial neural networks have proved effective in pattern recognition, one is the java pegon pattern that has its own uniqueness. One of the methods commonly used is Learning Vector Quantization of artificial neural network, this method is very good of classifying many patterns. Based on it, will be developed an application by implementing Artificial Neural Network which will apply 5 aspects of human brain performance capability in recognizing Java Pegon Letters. This study uses 160 image data, divided into 100 training data consisting of 5 normal image for each character and 60 test data consisting of 1 normal data, 1 corrupt data, and 1 noise data for each character. The data obtained from the processed capture so all data can have the same dimensions and size, ie length of 100 pixels and width of 100 pixels. All data will be processed through preprocessing and feature extraction. The results of feature extraction will be used in the training process to recognize the pattern by using Learning Vector Quantization of artificial neural network. The results of this study, Applications can apply 5 aspects of human brain performance capabilities very well, this is evidenced by the application's ability to recognize 100% of training data and test data. This application also has the ability to recognize abnormal data very well, such as data with interference or incomplete data.

Share this knowledge with your friends :

Actions (login required)

View Item