Cost-Sensitive Laplacian Logistic Regression for Ship Detention Prediction

Port state control (PSC) is the last line of defense for substandard ships. During a PSC inspection, ship detention most severe result if inspected identified with critical deficiencies. Regarding development prediction models, this paper identifies two challenges: learning from imbalanced data and unlabeled data. The first challenge, data, arises fact that minority ships were detained. second in practice not all foreign visiting receive formal leading to missing problem. To address these challenges, adopts machine paradigms: cost-sensitive semi-supervised learning. Accordingly, we expand traditional logistic regression (LR) model by introducing cost parameter consider different misclassification costs unbalanced classes incorporating graph regularization term Finally, conduct extensive computational experiments verify superiority developed framework paper. Computational results show into LR can improve classification rate almost 10%. In addition, considering models increase majority 1.33% 5.93%, respectively.