Search

Oil spill incidents can significantly impact marine ecosystems in Arctic/subarctic areas. Low biodegradation rate, harsh environments, remoteness, and lack of sufficient response infrastructure make those cold waters more susceptible to the impacts of oil spills. A major challenge in Arctic/subarctic areas is to timely select suitable oil spill response methods (OSRMs), concerning the process complexity and insufficient data for decision analysis. In this study, we used various regression-based machine learning techniques, including artificial neural networks (ANNs), Gaussian process regression (GPR), and support vector regression, to develop decision-support models for OSRM selection. Using a small hypothetical oil spill dataset, the modelling performance was thoroughly compared to find techniques working well under data constraints. The regression-based machine learning models were also compared with integrated and optimized fuzzy decision trees models (OFDTs) previously developed by the authors. OFDTs and GPR outperformed other techniques considering prediction power (> 30 % accuracy enhancement). Also, the use of the Bayesian regularization algorithm enhanced the performance of ANNs by reducing their sensitivity to the size of the training dataset (e.g., 29 % accuracy enhancement compared to an unregularized ANN). [Display omitted] • Machine learning-based models are developed for effective response selection. • Models' performance is compared on a hypothetical oil spill dataset in the Arctic. • The best-performing models suited for small training datasets are selected. • Optimized fuzzy decision trees and Gaussian process regression models outperformed. • Recommendations are provided to prevent models' overtraining on small datasets. [ABSTRACT FROM AUTHOR]

The challenging oil spill response in the Arctic calls for effective response decision support tools. In this study, a framework comprising the development of various integrated fuzzy decision tree and regression (FDTR) models as well as model optimization was developed to facilitate the selection of suitable response methods for oil spill accidents in Arctic waters. The FDTR models took into account the influential attributes affecting the effectiveness of oil spill response in harsh Arctic environments. Different FDTR models were developed based on the combinations of three regression analyses, including linear, non-linear, and Gaussian process regression (GPR) and four information evaluation measures for splitting a decision tree, including information gain, deviance, GINI impurities (GINI), and misclassification error. Non-dominated sorting differential evolution (NSDE) optimization was employed to enhance the predictive performance of the FDTR models. The prediction performance of the FDTR models was compared using an oil spill dataset. Using this framework, the average prediction accuracy and the number of rules (representing the robustness) of FDTRs were increased by 14% and decreased by 57%, respectively. A set of optimal prediction models to promptly select an appropriate response method can be obtained using this framework. Among all models, GPR-GINI performed the best concerning optimal values of objective functions. • Fuzzy decision trees (FDTs) were optimized for selecting oil spill response in the Arctic. • Multi-objective metaheuristic optimization improved the training of FDTs. • Different regression analyses were coupled with FDTs to improve prediction accuracy. • Four information discrimination measures were evaluated to construct efficient FDTs. [ABSTRACT FROM AUTHOR]

A fuzzy decision tree (FDT) based framework was developed to facilitate the selection of suitable oil spill response methods in the Arctic. Hypothetical oil spill cases were developed based on six identified attributes, while the suitability of three spill response methods (mechanical containment and recovery, use of chemical dispersants, and in-situ burning) for each spill case was obtained based on expert judgments. Fuzzy sets were used to address the associated uncertainties, and FDTs were then developed through generating: i) one decision tree for all three response methods (FDT-AP1) and ii) one decision tree for each response method and the development of linear regression models at terminal nodes (FDT-LR). The FDT-LR approach exhibited higher prediction accuracy than the FDT-AP1 approach. A maximum of 100% accurate predictions could be achieved for testing cases using it. On average, 75% of suitable oil spill response methods out of 10,000 performed iterations were predicted correctly. Unlabelled Image • A fuzzy decision tree-based tool is developed for selecting oil spill response method. • Fuzzy Interactive Dichotomizer 3 algorithm is used for developing decision trees. • The developed fuzzy decision trees are enhanced by linear regression models. • The maximum prediction accuracy of suitable selection by the tool can reach 100%. • The average prediction accuracy of suitable selection is 75% among 10,000 iterations. [ABSTRACT FROM AUTHOR]