PREDIKSI CURAH HUJAN DI DAERAH BANDARA SULTAN MAHMUD BADARUDDIN II MENGGUNAKAN JARINGAN SYARAF TIRUAN DENGAN METODE BACKPROPAGATION DAN RADIAL BASIS FUNCTION NETWORK
Daftar Isi:
- Indonesia is a tropical country with a high rainfall intensity. In this study was done the prediction of rainfall using Artificial Neural Network technique with Backpropagation's method and Radial Basis Function Network`s method. This prediction of rainfall was using data around the area of Sultan Mahmud Badaruddin II For prediction of rainfall number using backpropagation with neuron layer 25, in training phase produced MSE about 0.000000646. and in validation phase produced R. correlation about 0.72 and error reached 134.70% And prediction phase has result of negative value. for while prediction of rainfall number using Radial Basis Function Network with neuron 1ayer 5, in training phase produced MSE about 0.011869, and Jn validation phase produced R correlation about 0.90 and error reached 42,23%, And prediction phase has result of only positive value. Using Radial Basis. Function Network is better than Backpropagation method.