Summary of A Maritime Industry Experience For Vessel Operational Anomaly Detection: Utilizing Deep Learning Augmented with Lightweight Interpretable Models, by Mahshid Helali Moghadam et al.
A Maritime Industry Experience for Vessel Operational Anomaly Detection: Utilizing Deep Learning Augmented with Lightweight Interpretable Models
by Mahshid Helali Moghadam, Mateusz Rzymowski, Lukasz Kulas
First submitted to arxiv on: 30 Dec 2023
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The study presents a vessel operational anomaly detection approach that utilizes semi-supervised deep learning models augmented with lightweight interpretable surrogate models. The approach is applied to an industrial sensorized vessel called TUCANA and leverages standard and Long Short-Term Memory (LSTM) autoencoders trained on normal operational data and tested with real anomaly-revealing data. The inference results are projected onto a lower-dimension data map generated by t-distributed stochastic neighbor embedding (t-SNE), which serves as an unsupervised baseline and shows the distribution of identified anomalies. Additionally, lightweight surrogate models using random forest and decision tree are developed to promote transparency and interpretability for the inference results of the deep learning models and assist engineers with agile assessment of flagged anomalies. The approach is empirically evaluated using real data from TUCANA, demonstrating higher performance of the LSTM autoencoder as an anomaly detection module with effective capturing of temporal dependencies in the data. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper presents a way to find unusual events on a special vessel called TUCANA that uses sensors to collect data. They use a combination of deep learning models and simpler models that are easy to understand, which helps engineers make better decisions. The deep learning models learn from normal data and then identify unusual patterns in the real-world data. This is tested using actual data from TUCANA. The results show that one type of model does a great job of finding anomalies and understanding why they happen. |
Keywords
* Artificial intelligence * Anomaly detection * Autoencoder * Decision tree * Deep learning * Embedding * Inference * Lstm * Random forest * Semi supervised * Unsupervised
