Air temperature forecasting is important role in agriculture, flight, trading and so on. The method used for forecasting is Neural Network (NN). NN works as human neural system. One of NN type used for forecasting is Backpropagation where Backpropagation model is there are hidden layers between input and output. Due to forecasting result depends on data correlation, then this research will explain about the effect of data correlation on air temperature forecasting. To obtain forecasting result, Backpropagation algorithm will be used. Simulations are applied in three dataset with different structures. Based on simulation results, data which have strong correlation can result better forecasting based on smaller Mean Square Error (MSE).
Copyrights © 2018