IMPLEMENTASI JARINGAN SYARAF TIRUAN RECURRENT DENGAN METODE PEMBELAJARAN GRADIENT DESCENT ADAPTIVE LEARNING RATE UNTUK PENDUGAAN CURAH HUJAN BERDASARKAN PEUBAH ENSO
Journal: Comtech (Vol.1, No. 2)Publication Date: 2010-12-29
Authors : Afan Galih Salman; Yen Lina Prasetio;
Page : 418-429
Keywords : Artificial Neural Network (ANN) Recurrent Elman; ENSO; coefficient of determination (R2); Root Mean Square Error (RMSE); Gradient Descent Adaptive Learning Rate.;
Abstract
The use of technology of technology Artificial Neural Network (ANN) in prediction of rainfall can be done using the learning approach. ANN prediction accuracy measured by the coefficient of determination (R2) and Root Mean Square Error (RMSE).This research employ a recurrent optimized heuristic Artificial Neural Network (ANN) Recurrent Elman gradient descent adaptive learning rate approach using El-Nino Southern Oscillation (ENSO) variable, namely Wind, Southern Oscillation Index (SOI), Sea Surface Temperatur (SST) dan Outgoing Long Wave Radiation (OLR) to forecast regional monthly rainfall. The patterns of input data affect the performance of Recurrent Elman neural network in estimation process. The first data group that is 75% training data and 25% testing data produce the maximum R2 69.2% at leap 0 while the second data group that is 50% training data & 50% testing data produce the maximum R2 53.6%.at leap 0 Our result on leap 0 is better than leap 1,2 or 3.
Other Latest Articles
- ANALISIS DAN PERANCANGAN SISTEM INFORMASI AKUNTANSI PENJUALAN JASA DESAIN
- AN INTRODUCTION TO DIGITAL CASH WORLD
- PEMANFAATAN APLIKASI E-MARKETING PADA PT JEAH UNTUK MEMENANGKAN PERSAINGAN USAHA
- PEMANFAATAN WEB APPLICATION DALAM PENINGKATAN EFISIENSI DAN EFEKTIFITAS KERJA ADMINISTRASI (STUDI KASUS: PEMESANAN DAN PEMBUATAN SURAT KETERANGAN)
- SISTEM PENUNJANG KEPUTUSAN UNTUK PENANGANAN PENDERITA GAGAL JANTUNG
Last modified: 2015-11-17 15:52:20