A Study on Price Prediction of Used Sailboats Based on XG-Boost Algorithm

Abstract

The sailing industry is currently receiving more and more attention and there is a growing demand for sailboats. Their prices are constantly changing and inevitably contain useful trends and intrinsic correlations. Therefore, we developed a used sailboat price prediction model to allow relevant practitioners to better understand the sailboat market. First, we obtained some key data from the official websites of various sailboat manufacturers that may affect the price, including width, draft, displacement, sail area, headroom, and so on. We pre-processed the data. Next, based on the XG-Boost algorithm, we built a used sailboat pricing prediction model and estimated its accuracy, which reached about 91.5% in the test set. Finally, we did a one-way ANOVA on region and sailboat prices and found that there are significant differences in sailboat prices across regions. Therefore, we conducted a consistency test for region effects and found that the region effects for displacement and sail area were consistent.