A Novel Evaluation Metric for Unsupervised Learning in AIS-Based Maritime Anomaly Detection: MADQI - Dataset and Python Code
Fuente:
Zenodo
Saved in:
| Main Author: | |
|---|---|
| Format: | Recurso digital |
| Language: | English |
| Published: |
Zenodo
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866902246500859904 |
|---|---|
| author | Gocer, Ismet |
| author_facet | Gocer, Ismet |
| contents | <p>This repository contains the dataset, Python notebooks, and supporting materials used in the study:</p> <p><strong>“A Novel Evaluation Metric for Unsupervised Learning in AIS-Based Maritime Anomaly Detection: MADQI”</strong></p> <p>The study proposes a novel evaluation framework named the <strong>Maritime Anomaly Detection Quality Index (MADQI)</strong> for assessing unsupervised anomaly detection performance in Automatic Identification System (AIS)-based maritime surveillance applications without requiring labelled data.</p> <p>The repository includes:</p> <ul> <li>AIS-based maritime trajectory dataset used in the experiments</li> <li>Python notebooks for data preprocessing, feature engineering, anomaly detection, and evaluation</li> <li>Haversine distance-based spatial analysis implementation</li> <li>Isolation Forest-based unsupervised anomaly detection workflow</li> <li>MADQI metric calculation modules</li> <li>Visualisation and analysis scripts</li> <li>Example outputs and evaluation results</li> </ul> <p>The proposed MADQI framework integrates four complementary evaluation components:</p> <ul> <li><strong>ARC</strong>: Anomaly Rate Consistency</li> <li><strong>PPS</strong>: Physical Plausibility Score</li> <li><strong>SDS</strong>: Score Distribution Separation</li> <li><strong>ECE</strong>: Extreme Case Evidence</li> </ul> <p>The notebooks demonstrate the complete experimental pipeline, including:</p> <ol> <li>AIS data preprocessing</li> <li>Temporal and spatial feature extraction</li> <li>Vessel movement analysis</li> <li>Anomaly detection using unsupervised learning</li> <li>Multi-component MADQI evaluation</li> <li>Quantitative and visual interpretation of anomaly behaviour</li> </ol> <p>The implementation is designed to support reproducibility, transparency, and further research in maritime anomaly detection, unsupervised learning evaluation, and AI-driven maritime surveillance systems.</p> <h3>Keywords</h3> <p>AIS, Maritime Anomaly Detection, Unsupervised Learning, Isolation Forest, MADQI, Maritime AI, Machine Learning, Haversine Distance, Vessel Behaviour Analysis, Maritime Surveillance, Explainable AI, Anomaly Detection Evaluation</p> <h3>Author</h3> <p><span class="hover:entity-accent entity-underline inline cursor-pointer align-baseline"><span class="whitespace-normal">Ismet Gocer</span></span></p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_20425539 |
| institution | Zenodo |
| language | eng |
| publishDate | 2026 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | A Novel Evaluation Metric for Unsupervised Learning in AIS-Based Maritime Anomaly Detection: MADQI - Dataset and Python Code Gocer, Ismet AIS Maritime Securtiy Ship Tracking Anomaly detection New Evaluation Metric for Unsupervised Learning Machine Learning <p>This repository contains the dataset, Python notebooks, and supporting materials used in the study:</p> <p><strong>“A Novel Evaluation Metric for Unsupervised Learning in AIS-Based Maritime Anomaly Detection: MADQI”</strong></p> <p>The study proposes a novel evaluation framework named the <strong>Maritime Anomaly Detection Quality Index (MADQI)</strong> for assessing unsupervised anomaly detection performance in Automatic Identification System (AIS)-based maritime surveillance applications without requiring labelled data.</p> <p>The repository includes:</p> <ul> <li>AIS-based maritime trajectory dataset used in the experiments</li> <li>Python notebooks for data preprocessing, feature engineering, anomaly detection, and evaluation</li> <li>Haversine distance-based spatial analysis implementation</li> <li>Isolation Forest-based unsupervised anomaly detection workflow</li> <li>MADQI metric calculation modules</li> <li>Visualisation and analysis scripts</li> <li>Example outputs and evaluation results</li> </ul> <p>The proposed MADQI framework integrates four complementary evaluation components:</p> <ul> <li><strong>ARC</strong>: Anomaly Rate Consistency</li> <li><strong>PPS</strong>: Physical Plausibility Score</li> <li><strong>SDS</strong>: Score Distribution Separation</li> <li><strong>ECE</strong>: Extreme Case Evidence</li> </ul> <p>The notebooks demonstrate the complete experimental pipeline, including:</p> <ol> <li>AIS data preprocessing</li> <li>Temporal and spatial feature extraction</li> <li>Vessel movement analysis</li> <li>Anomaly detection using unsupervised learning</li> <li>Multi-component MADQI evaluation</li> <li>Quantitative and visual interpretation of anomaly behaviour</li> </ol> <p>The implementation is designed to support reproducibility, transparency, and further research in maritime anomaly detection, unsupervised learning evaluation, and AI-driven maritime surveillance systems.</p> <h3>Keywords</h3> <p>AIS, Maritime Anomaly Detection, Unsupervised Learning, Isolation Forest, MADQI, Maritime AI, Machine Learning, Haversine Distance, Vessel Behaviour Analysis, Maritime Surveillance, Explainable AI, Anomaly Detection Evaluation</p> <h3>Author</h3> <p><span class="hover:entity-accent entity-underline inline cursor-pointer align-baseline"><span class="whitespace-normal">Ismet Gocer</span></span></p> |
| title | A Novel Evaluation Metric for Unsupervised Learning in AIS-Based Maritime Anomaly Detection: MADQI - Dataset and Python Code |
| topic | AIS Maritime Securtiy Ship Tracking Anomaly detection New Evaluation Metric for Unsupervised Learning Machine Learning |
| url | https://doi.org/10.5281/zenodo.20425539 |