Data analysis based on the second-hand sailing market

Abstract

As the second-hand sailboat market continues to develop, the value of sailboats changes with their aging and changes in market conditions. This article mainly analyzed and predicted the second-hand sailing market. To predict and evaluate the changes in the second-hand sailboat market, we exclude human factors such as policies, technology, and environment. To build a model for the second-hand sailboat market, we first clean the data by removing duplicates and irrelevant information. We then extract feature variables and use Pearson correlation analysis to filter out irrelevant or low-correlation feature variables. Next, we construct an XGBoost+SHAP model to predict the impact of regions on the second-hand sailboat market, as well as the impact of regions on single and double sailboats, and perform comparative analysis. To highlight the effectiveness of the training and validate the model, we input the data into various models such as Random Forest, Decision Tree, and Logistic Regression. We find that the XGBoost+SHAP algorithm model has the highest accuracy. Finally, we evaluate the contribution of each feature variable to the predictive power by measuring the positive and negative correlations between individual sailboat variables and regions, revealing the impact of feature variables on second-hand sailboat prices under different values.