Piracy Attacks on Ships in Navigation Through High Risk Area (HRA): Prediction Use Regression Models

The objective of this study was to create a mathematical model of the threat frequency of ships subjected to pirate attacks HRA. For this purpose, data on successful pirate attacks from 2011-2020 were observed and analysed.Regression analysis aims to express or describe the relationship between the observed phenomena, the corresponding analytical - mathematical expressions, or regression model. Such a model, except to explain addiction phenomena, to predict the value of the dependent variable for the standard value of an independent or explanatory variable, depending on whether it is a simple or multiple regression in the regression model.Using linear and various forms of nonlinear regression, five models were calculated and used for prediction of possible successful pirate attacks for years 2021-2029. The models clearly confirm the decreasing trend of successful pirate attacks for ships most affected by pirate attacks (bulk carriers, crude oil tankers, passenger ships and other medium-sized vessels) and predict their further decline and possible disappearance. In addition, the one-way measure replication ANOVA together with the Bonfferoni correction revealed significant(p<0.05)differences in the extrapolated values between the high-oscillatory model and the linear and cubic models