Using A Time Series Forecasting Model, Predict Maritime Crime

Abstract

Indonesia is very disadvantaged by the many illegal activities on the sea route. One of the most vulnerable areas in Indonesia is the Riau Islands. This study aims to analyze marine crimes in the Riau Islands Province. For this reason, we ask a research question: How to predict the number of cases of marine crimes in the future along with the magnitude of the risk posed and the necessary handling? To answer, we designed mixed-method research. We conducted a literature study, collected data from agencies that handle sea transportation, customs, patrols, and law enforcement in the Riau Islands maritime area, and analyzed the data obtained. These results are used to fill out AHP forms and interviews to determine the magnitude of the risk posed by each type of marine crime. Finally, we forecast and analyze the results to formulate conclusions and recommendations. In this study, we use the ARIMA model to predict the number of marine crimes incidents in the future and descriptive analysis to determine the resulting risk classification. The findings of this study indicate that in the next year, it is predicted that there will be 91 events with the highest incidence in April, namely 11 events with a relative risk of 6.3. These findings imply that the Riau Islands Region is an area that is vulnerable to severe marine crimes. Therefore, it is recommended to take measures that anticipate the occurrence of violations early, such as the installation of warning sensor equipment and other adequate sensing technologies.