本研究旨在分析2018~2022年QS世界大学排名前350名高校的数据,以揭示全球高等教育的发展趋势。采用统计模型和集成树模型,结合数据挖掘与可视化分析,对高校数量、城市分布、QS得分及排名进行了深入研究。研究发现,美国连续五年位居上榜...本研究旨在分析2018~2022年QS世界大学排名前350名高校的数据,以揭示全球高等教育的发展趋势。采用统计模型和集成树模型,结合数据挖掘与可视化分析,对高校数量、城市分布、QS得分及排名进行了深入研究。研究发现,美国连续五年位居上榜高校数量榜首,伦敦和纽约在高校数量上占据优势。通过比较不同回归模型和集成树模型,XGBoost算法在QS得分和排名预测中表现最优。本文还通过可视化分析评估了QS排行榜的稳定性和一致性。研究结果为高等教育发展提供参考,并为高校管理者和决策者提供决策支持。This study aims to analyze the data of the top 350 universities in the QS World University Rankings from 2018 to 2022 to reveal the development trends of global higher education. Statistical models and ensemble tree models, combined with data mining and visualization analysis, were used to conduct an in-depth study on the number of universities, city distribution, QS scores, and rankings. The study found that the United States has been at the top of the list for five consecutive years in terms of the number of universities listed, and London and New York have an advantage in the number of universities. By comparing different regression models and ensemble tree models, the XGBoost algorithm performed best in predicting QS scores and rankings. This paper also evaluates the stability and consistency of the QS rankings through visualization analysis. The findings provide a reference for the development of higher education and decision support for university administrators and policymakers.展开更多
文摘本研究旨在分析2018~2022年QS世界大学排名前350名高校的数据,以揭示全球高等教育的发展趋势。采用统计模型和集成树模型,结合数据挖掘与可视化分析,对高校数量、城市分布、QS得分及排名进行了深入研究。研究发现,美国连续五年位居上榜高校数量榜首,伦敦和纽约在高校数量上占据优势。通过比较不同回归模型和集成树模型,XGBoost算法在QS得分和排名预测中表现最优。本文还通过可视化分析评估了QS排行榜的稳定性和一致性。研究结果为高等教育发展提供参考,并为高校管理者和决策者提供决策支持。This study aims to analyze the data of the top 350 universities in the QS World University Rankings from 2018 to 2022 to reveal the development trends of global higher education. Statistical models and ensemble tree models, combined with data mining and visualization analysis, were used to conduct an in-depth study on the number of universities, city distribution, QS scores, and rankings. The study found that the United States has been at the top of the list for five consecutive years in terms of the number of universities listed, and London and New York have an advantage in the number of universities. By comparing different regression models and ensemble tree models, the XGBoost algorithm performed best in predicting QS scores and rankings. This paper also evaluates the stability and consistency of the QS rankings through visualization analysis. The findings provide a reference for the development of higher education and decision support for university administrators and policymakers.
