Vessel Valuation Based On BP Neural Network

Vessels are the fundamental elements of shipping market. With thequick development of China shipping industry and more frequentshipping related activities, vessel valuation becomes more and moreimportant. The service of vessel valuation has become important supportfor sales, chartering, mortgage, insurance and auction of ships.In order to get accurate ship price, systematic analysis andclassification of factors influencing ship price are made. According todifferent types and ages of vessels, factors are selected. Three methods ofvessel valuation have been analyzed and compared with each other.In this paper, BP neural network is used to calculate the ship pricefor its significant capacity to non-linearity simulate and grasp the inherentlaws of problems by self-study. By applying the transaction informationof297deals of handy-size bulk carriers from2009to2012from Clarkson,two vessel valuation models based on BP neural network are established,one model with ship’s age less than20years concerning factors of age,DWT, new-building price, one-year time charter rate, speed and bunkerconsumption of vessel, the other with ship’s age not less than20yearsconcerning factors of age, DWT, demolition price and one-year time charter rate.The results of both networks indicate the rationality of the method,with relative error rates of output of networks and actual prices of the twomodels within the limits of accepting scope.