Trip generation and attraction model and forecasting using machine learning methods

Abstract

The necessity of people and goods to be moved from one place to another place has increased dramatically in recent years. It requires numbers of connectivity among the regions. Besides, policy changes in sea transportation sector including development of sea transport infrastructures as well as shipping/ferry lines to support fulfilling these needs are introduced. The ferry line from Kuala Langsa Port, Indonesia to Penang Port, Malaysia was introduced to encounter the need of mobility between those regions. In consequence, it is important to estimate future transport demand. This study is conducted to familiarize the use of machine learning methods in modelling and forecasting trip generation and trip attraction. Time-series trip generation and attraction data from Kuala Langsa to Penang and vice versa and socio-economic data were employed to develop the model. The result shows that gross domestic regional product (GDRP) and population variables has significant influence to generate trips between these ports.