Machine Learning Approaches to Maritime Anomaly Detection/Postupci strojnog ucenja za otkrivanje pomorskih anomalija

Topics related to safety in maritime transport have become very important over the past decades due to numerous maritime problems putting both human lives and the environment in danger. Recent advances in surveillance technology and the need for better sea traffic protection led to development of au...

Full description

Saved in:
Bibliographic Details
Published in:Naše more znanstveni časopis za more i pomorstvo 2014-01, Vol.62 (5-6), p.96-96
Main Authors: Obradovic, Ines, Milicevic, Mario, Zubrinic, Krunoslav
Format: Article
Language:English
Subjects:
Anomalies
Human
Machine learning
Maritime
Mathematical models
Normality
Surveillance
Transport
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Topics related to safety in maritime transport have become very important over the past decades due to numerous maritime problems putting both human lives and the environment in danger. Recent advances in surveillance technology and the need for better sea traffic protection led to development of automated solutions for detecting anomalies. These solutions are based on generating normality models from data gathered on vessel movement, mostly from AIS. This paper provides a presentation of various machine learning approaches for anomaly detection in the maritime domain. It also addresses potential problems and challenges that could get in the way of successful automation of such systems.
ISSN:0469-6255
1848-6320