Prediksi Kecepatan Arus Laut dengan Menggunkan Metode Backpropagation (Studi Kasus: Labuhan Bajo)
Keywords:Labuhan Bajo; Sea Current Speed; Prediction; Backpropagation; Neural Network
One factor that is very influential on the dynamics of the waters is the speed of ocean currents. The speed of the ocean currents has an impact on activities around the coast that is for tourists to get information about the condition of the movement of the sea. One of them is in Labuhan Bajo. Labuhan Bajo is a tourist area that has a variety of natural beauty where visitors increase every year. The influence of the west monsoon wind in Labuan Bajo is very large on the condition of sea movement, especially on ocean currents. Predictions about the speed of ocean currents are very important in marine activities, especially diving because it is an effort to prevent the occurrence of things that are not desirable because of the condition of the sea that is not conducive. In this study the method used in predicting the current speed is the Backpropagation method. By testing the hidden layer nodes and the learning rate on the Backpropagation method the best MAPE results are obtained from sharing 70% of training data with 100 hidden layer nodes and the learning rate of 0.1 is 7.59%. Whereas by sharing 80% of the best MAPE training data, there are 100 hidden layer nodes and the learning rate of 0.1 is 0.57%. Then from 90% of the data sharing training data obtained the best MAPE results in the hidden layer node 100 and a learning rate of 0.4 out of 6.65%, this shows that the Backpropagation method is very well used in predicting the speed of ocean currents.