An ensemble method for investigating maritime casualties resulting in pollution occurrence: Data augmentation and feature analysis

•Prediction of maritime casualties resulting in pollution occurrence powered by AI technology.•Employment of VAE based data augmentation to address data imbalance challenge.•Utilization of up-to-date maritime casualty data with various pollution sources.•Analysis of contributing features and their d...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Reliability engineering & system safety Ročník 251; s. 110391
Hlavní autoři: Li, Duowei, Wong, Yiik Diew, Chen, Tianyi, Wang, Nanxi, Yuen, Kum Fai
Médium: Journal Article
Jazyk:angličtina
Vydáno: Elsevier Ltd 01.11.2024
Témata:
Data augmentation
Machine learning
Maritime casualty
Maritime pollution
Pollution prediction
Variational autoencoder (VAE)
Maritime casualty
Variational autoencoder (VAE)
Data augmentation
Maritime pollution
Pollution prediction
Machine learning
ISSN:0951-8320
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:•Prediction of maritime casualties resulting in pollution occurrence powered by AI technology.•Employment of VAE based data augmentation to address data imbalance challenge.•Utilization of up-to-date maritime casualty data with various pollution sources.•Analysis of contributing features and their dependences on pollution occurrence.•Discussion of practical insights for precautionary measures and policy development. Timely prediction of maritime casualties resulting in pollution occurrence remains unsolved in academia, as the significant data imbalance between non-polluting and polluting casualties poses a challenge to prediction efficacy. This study proposes an ensemble method for predicting polluting maritime casualties and exploring the contributing features to pollution. In the data preprocessing phase, key features related to casualties and vessels are extracted and encoded into model variables; in the data augmentation phase, Variational Autoencoder is employed to generate synthetic samples from the minor class, effectively mitigating the impact from data imbalance; and in the pollution indicator classification phase, machine learning models are trained on the balanced dataset to label a casualty as “polluting” or “non-polluting”. A dataset containing 25,414 worldwide maritime casualties from 2013 to 2023 is utilized for method validation. Several state-of-the-art data balancing techniques serve as baselines for comparison with the VAE on the quality of generated synthetic data. The model trained on the VAE dataset achieves the most satisfactory performances, demonstrating the superiority of VAE in augmenting data quantity and diversity. “Casualty cause”, “Vessel age” and “Vessel type” are revealed as the top three contributing features to pollution. Several insights are discussed for precautionary measures and policy development.
ISSN:0951-8320
DOI:10.1016/j.ress.2024.110391