<![CDATA[Weather is an atmospheric condition that occurs in a narrow area with a short space of time. Observations of weather elements are needed in everyday life. for it can affect the safety of air transportation. The weather element that is often predicted is rainfall. Rainfall in tropical regions such as Indonesia is one of the parameters that can describe weather conditions in general. The method used to predict rainfall was artificial neural network with backpropagation algorithm. The purpose of this paper is to apply the model of artificial neural networks with backpropagation algorithm to predict daily rainfall and to determine prediction accuracy based on Mean Square Error (MSE). The network used has 3 layers namely input layer. hidden layer. and output layer with 7 input neurons, 12 hidden neurons, and 1 output neuron. The activation function used were bipolar sigmoid function and linear function. Based on data analysis carried out using network architecture and parameters that had been determined with 578 data at the training stage. MSE values of 25,0639 was obtained and based on the results of the network training process. the prediction was quite well. In the testing stage. the model developed using data as much as 145. MSE value of 405,1994 was obtained. MSE obtained during the testing stage was greater than that of obtained during the training process due to several factors. one of them is because the weather is volatile so the weather conditions vary every year and global warming causes weather conditions to be unpredictable.]]>
