Jaringan Syaraf Tiruan Model Backpropagation untuk Peramalan Suhu Minimum dan Maksimum, Kelembaban, Tekanan Udara, Jumlah Hari Hujan, dan Curah Hujan Bulanan di Kota Mataram

Abstract

Climate conditions are very influential on various sectors of human life, such as agriculture. Extreme climate can affect the planting and harvest period which can eventually lead to crop failure. Climate forecasting can be done to help mapping the planting period so that agricultural productivity is optimal. Artificial neural network is a method that can be used to forecast climate in the future. In this research, a backpropagation model was used to forecast monthly data on minimum and maximum temperatures, humidity, air pressure, number of rainy days, and rainfall in Mataram city. The accuracy of forecasting result is measured based on MSE (mean square error). Based on the research conducted, the MSE of testing phase are minimum temperature 3.59 x 10^-3, maximum temperature 1.97 x 10^-4, humidity 2.11 x 10^-3, air pressure 2.50 x 10^-3, number of rainy days 7.97 x 10^-4, and rainfall 1.00 x 10^-3.