IDENTIFIKASI BENTROKAN ANTAR PESERTA KAMPANYE MENGGUNAKAN METODE BACKPROPAGATION

Abstract

Abstract –This study is focused on uniform color used during the election campaign of 2014. Searches color difference under certain conditions using Artificial Neural Network algorithm and backpropagation method. Computational methods based Neural Network algorithm is hoped able to identify clashes in a uniform color in the campaign object image and analyzing and comparing the success rate of the algorithm.
Of object images used in this study are a variety of objects moving or static color taken in the form of a video that ultimately framing and finishing into a moving image to the next sampled include the election campaign participants crowd situation in Semarang on 2014. From the results of all the obtained yield 89.3% accuracy rate.
Artificial Neural Networks is one of the information or data processing system that is designed to mimic the way the human brain works in solving the problem by changing the weights sinapsision.

Abstrak–Penelitian ini dipusatkan perhatian pada warna seragam yang dipakai saat kampanye pemilu 2014. Pencarian perbedaan warna dalam kondisi tertentu ini menggunakan algoritma Jaringan Syaraf Tiruan dan metode Backpropagation. Metode komputasi berbasis algoritma Jaringan Syaraf Tiruan Diharapkan mampu mengidentifikasi bentrokan pada warna seragam kampanye didalam objek image dan menganalisis dan membandingkan tingkat keberhasilan algoritma tersebut.

Dari hasil semua yang diperoleh menghasilkan nilai akurasi 89.3%.

Jaringan Syaraf Tiruan merupakan salah satu sistem pemrosesan informasi atau data yang didesain untuk menirukan cara kerja otak manusia dalam menyelesaikan masalah melalui perubahan bobot sinapsisnya.