IMPLEMENTASI TEKNIK REGRESI PADA DATA PENUMPANG BUS MENGGUNAKAN NEURAL NETWORK REGRESSION (NNR) (STUDI KASUS PT. ROSALIA INDAH TRANSPORT SURABAYA)

In the world of transportation, bus vehicles are no stranger to the community. Bus transportation is still the main choice for people who want to travel far. PT. Rosalia Indah Transport Surabaya is a means of transportation of land vehicles in the form of buses. With 5 departments namely Surabaya - Palembang, Surabaya - Bitung, Surabaya - Bogor, Surabaya Ciputat and Surabaya - Merak, every day Rosalia Indah bus dispatch 5 buses for 5 departments. But when the big days and holidays the number of passengers jumps high and can not be predicted. For the convenience of passengers in transportation, it is necessary a system that can predict the number of passengers in the future so that the Rosalia can prepare an additional bus with the right. To predict data the number of bus passengers in this study used Neural Network. Type of Neural Network used is Multilayer Perceptron. Regression on the number of bus passengers produced predictions with the highest error values of 0,089 and 0,045 measured using Root Mean Squarred Error (RMSE) and Mean Absolute Error (MAE) for the 5 existing majors. The result of testing shows that Multilayer Perceptron on Neural Network can be used in data regression of passenger bus number of PT. Rosalia Indah Transport Surabaya