泰坦尼克号生存预测模型的设计

Design of Titanic survival prediction model

针对目前海上船舶事故,自然灾害等救援效果不理想,生存预测较难的情况,文中通过设计决策树模型、AdaBoost模型、随机森林预测模型对泰坦尼克号游客生存进行预测。首先对泰坦尼克号中乘客的各个特征进行数据收集,然后通过三种预测模型实验对比分析得出随机森林模型的精度和准确度都较高,且不易出现过拟合和欠拟合现象。相关生存预测模型已成功运用于海上船舶事故等自然灾害人员生存预测当中,实验验证了相关模型的可靠性与有效性。

Based on the current situation of unsatisfactory rescue effects such as natural disasters such as marine accidents and the suffering of survival prediction,this paper uses the decision tree model,AdaBoost model,and random forest prediction model to predict the survival of Titanic tourists.Firstly,the data on the various characteristics of passengers in the Titanic is collected,and then compare and analyze the three prediction model experiments to find that the precision and accuracy of the random forest model are high,and is not easy to appear over-fitting and under-fitting.The related survival prediction model has been successfully applied to the survival prediction of people from natural disasters such as marine ship accidents.Experiments have verified the high reliability and effectiveness of the related models.