IMPLEMENTASI JARINGAN SYARAF TIRUAN BACKPROPAGATION UNTUK PREDIKSI PENYEDIAAN ENERGI LISTRIK (STUDI KASUS PLN RAYON RUMBAI)
No Thumbnail Available
Date
2020-10
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This study applies the backpropagation algorithm to determine the supply of electrical
energy with six variables, namely the amount of electricity for social users, the amount
of electricity for household users, the amount of electricity for business users, the
amount of electricity for industrial users, the amount of electricity for government users
and the amount of electricity for special services. This study uses 12 data with the
division of training data and test data that is 90%: 10%, 80%: 20% and 70%: 30% of 12
data. This study also uses several parameters namely learning rate (α) 0.1 to 0.9,
maximum epoch 50 and six hidden neurons. The results of the test show the highest
average prediction accuracy of 99,90% at 90%: 10% data sharing, six hidden neurons,
learning rate (α) of 0.8 and the number of epoch 65 and testing of Mean Square Error
(MSE) of 0.00009.
Description
Keywords
Backpropagation, Electrical Energy, Artificial Neural Network, Mean Square Error, Prediction