Machine learning applications for risk assessment in maritime transport: Current status and future directions

Yuqing Lin, Xue Li, Kum Fai Yuen*

*Corresponding author for this work

Research output: Contribution to journalShort surveypeer-review

Abstract

As the maritime transportation system is inherently complex and vulnerable to potential hazards, it is critical to conduct risk assessments via applicable methodologies. Recently, machine learning (ML) algorithms have attracted tremendous attention due to their ability to analyze risks effectively. Nevertheless, there is a lack of a systematic summarization of ML applications in maritime transport risk assessment (MTRA). Hence, this review aims to encapsulate the current status, issues, considerations, and future directions of ML applications using the systematic reviews and meta-analyses method. In particular, the status is summarized from the following three dimensions: advantages, disadvantages, and corresponding applications. Moreover, several issues are recognized, including dataset processing and methods utilization, and considerations from the perspective of sensitivity analysis and evaluation methods. Regarding future directions, promising opportunities in terms of data and method improvements are identified. Overall, this review contributes to MTRA by presenting the existing research status with a framework and providing suggestions on model selection and method improvement for future research.

Original languageEnglish
Article number110959
JournalEngineering Applications of Artificial Intelligence
Volume155
DOIs
Publication statusPublished - Sept 1 2025
Externally publishedYes

Bibliographical note

Publisher Copyright:
© 2025

ASJC Scopus Subject Areas

  • Control and Systems Engineering
  • Artificial Intelligence
  • Electrical and Electronic Engineering

Keywords

  • Accident prediction
  • Machine learning
  • Maritime transport
  • Risk assessment
  • Supervised learning

Fingerprint

Dive into the research topics of 'Machine learning applications for risk assessment in maritime transport: Current status and future directions'. Together they form a unique fingerprint.

Cite this