A novel risk assessment system for port state control inspection

Abstract

Port State Control (PSC) inspection is the most important mechanism to ensure world marine safe. Recently, some SVM-based risk assessment systems have been presented in the world. They estimate the risk of each candidate ship based on its generic factors and history inspection factors to select high-risk one before conducting on-board PSC inspection. However, how to improve the performance of the PSC inspection under the situation of noisy data when applying SVM is still a challenging problem. In this paper, we propose a new approach for PSC inspection, which uses a novel Support Vector Machine and K-nearest neighbor (KNN-SVM) to remove noisy training examples and Bag of Words (BW) to extract some new target factors for the PSC inspection database. The experimental resultsshow that the generalization performance and the accuracy of risk assessment are improved significantly compared to that of the traditional SVM classifier, and adapt to engineering applications.

A bibliometric review on maritime inspection analysis: Current and future insights

“…Predictive analytics XGBoost model 55 Machine learning 52 Bayesian network 19,22,23,53,[56][57][58][59] Support vector machine 31,54,60,61 Binary logistic regression 21,62,63 Bayesian belief networks 64 Q-learning algorithm 65 Poisson regression 27 Blockchain 20 Markov chains 64 K-nearest neighbor 65 from ships and inspections are expanded on the basis of possible methods and approaches, it may be possible to concentrate the studies on the prescriptive analytic approach in the future perspectives.…”

An Artificial Intelligence Model Considering Data Imbalance for Ship Selection in Port State Control Based on Detention Probabilities

“…There is a large volume of published studies on PSC inspection. Several studies have investigated how to improve the efficiency and effectiveness of PSC inspection, including how to select ships for inspection (Xu et al, 2007a(Xu et al, , 2007bGao et al, 2008;Degré, 2007Degré, , 2008Yang et al, 2018aYang et al, , 2018bWang et al, 2019;Heij and Knapp, 2019;Knapp and Heij, 2020), how to assign PSCOs considering their background and expertise and the deficiency conditions of the ships , and how to determine the onboard inspection sequence (Tsou, 2018;. Besides, various studies have assessed the impact of PSC inspection on reducing incidents and accidents (Hänninen et al, 2014;Li et al, 2014;Fan et al, 2020), on future PSC inspection results (Cariou et al, 2008), on ship behavior (Cariou and Wolff, 2011;Fan et al, 2014), and on protecting the marine environment (Heij et al, 2011).…”

“…The system was further improved by combing support vector machine with K-nearest neighbor and bag-ofwords model (Gao et al, 2008). Numerical experiments of the three studies suggested that among all the ships predicted to be detained, only about 20% of them were actually detained.…”

“…To the best of our knowledge, there are three studies that aim to predict detention rate of all ship types visiting a port in current literature: Xu et al (2007a), Xu et al (2007b), andGao et al (2008). The three studies all use the precision of ships in class "1" (i.e.…”

“…Although ship detention is the most importance decision in PSC inspection and the detention rate is usually low, several steps in a PSC inspection, ranging from ship selection to ship inspection and final decision making, might lead to inaccuracy and inefficiency. First, several studies have reported that the SRP ship selection scheme is inefficient for targeting high-risk ships (Xu et al, 2007a(Xu et al, , 2007bGao et al, 2008;Wang et al, 2019). As a result, ships selected for inspection by the port states using SRP are not necessarily the ones with the largest number of deficiencies and highest probability of detention.…”