Design and Build a Passenger Prediction Information System Using Gated Recurrent Unit Deep Learning Algorithm

Sumenep is a district located in the eastern part of the island of Madura. Sumenep Regency has many islands that are still within its scope. Sea transportation is an essential means of traveling to Sapudi Island. Therefore, the increase in passengers needs to be considered in sailing safety. To make the prediction process for the number of sea transportation passengers, the Gated Recurrent Unit (GRU) method is used as a method in deep learning that is capable of making predictions. The data obtained is passenger data for the years 2020 - 2023 with a seasonal data type. The suitable parameters during the GRU method testing were the number of neurons, epochs, and learning rates of 64,200 and 0.1 with a MAPE error of 9.93. With an error value
$ < 10 \%$
, then GRU can be ascertained as an excellent method for predicting the amount of sea transportation passenger data in Sumenep towards Sapudi Island. To carry out the prediction process, the application is based on WEB so that all users can use and find out the results of passenger predictions using the GRU method.