Call Number | SEM-211 |
Collection Type | Indeks Artikel prosiding/Sem |
Title | JARINGAN SYARAF TIRUAN UNTUK IDENTIFIKASI KUALITAS PANEN DAN PENYAKIT IKAN PATIN MENGGUNAKAN ALGORITMA BACKPROPAGATION (STUDI KASUS DI DINAS KELAUTAN DAN PERIKANAN |
Author | Pariyadi; |
Publisher | Prosiding Senatkom |
Subject | Artifcial Neural Network (ANN), Backpropagation, Quality of Harvest Catfish, Pangasius Desease |
Location |
Nomor Panggil | ID Koleksi | Status |
---|---|---|
SEM-211 | TERSEDIA |
Freshwater fish farming is a promising opportunity that can improve people's income. One of the species of freshwater fish that is popular nowadays for cultivated is Pangasius. In the process of cultivation, Pangasius are vulnerable to attacked by micro- organisms that can cause a disease that affects the quality of Pangasius during harvest. So far, Pangasius farmers especially beginner farmers are constrained in identifying the Pangasius quality and Pangasius diseases. Pangasius quality in Jambi province is categorized into three, namely: Prima quality, Medium Quality and Low Quality which indicated attacked by a disease. 5 diseases that usually attack Pangasius in Jambi Province : Edwardsiella Ictaluri, Edwardsiella Tarda, Yersinia Ruckeri, Pseudomonas Anguilliseptica, dan Motil Aeromonas. Based on those problem, this study aims to applying Artificial Neural Network (ANN) model with Backpropagation algorithm in identifying harvest quality and Pangasius diseases. ANN is able to recognize the pattern of harvest quality and Pangasius diseases based on condition and clinical symptoms. The data will be trained and new knowledge that never been studied will be given. 15 variabels will be used as an input, and then will be trained by using 65 Pangasius data so that the system can recognize the variabel and the data properly. For testing process, 20 data will be used. After analysis of simulation ANN models with Backpropagation algorithm by using Matlab, the quality of harvest and Pangasius diseases will be identified. The results are Backpropagation algorithm is able to have 100% of accuracy level. Evaluation is done to obtain the best condition the number of neurons in the hidden layer of 20 nodes with 14 iterations, 0.00000000186 of MSE and the best training function is trainlm with 0.001671590778 average of error gradient. Keywords: Artifcial Neural Network (ANN), Backpropagation, Quality of Harvest Catfish, Pangasius Desease