本文以2022年成都的二手房房价数据为研究对象,构建随机森林模型和XGBoost模型来预测二手房价格。首先对数据集进行清洗并可视化处理,构建虚拟变量,接着绘制热力图并运用熵值法进行特征值筛选,选取重要的特征进行训练模型。接着,采用网...本文以2022年成都的二手房房价数据为研究对象,构建随机森林模型和XGBoost模型来预测二手房价格。首先对数据集进行清洗并可视化处理,构建虚拟变量,接着绘制热力图并运用熵值法进行特征值筛选,选取重要的特征进行训练模型。接着,采用网格搜索技术分别开发了基于随机森林和XGBoost的预测模型,并利用决定系数、均方误差和平均绝对误差这三个关键指标来衡量模型的预测准确性,经过模型比较和结果分析,发现优化后的XGBoost模型对二手房房价有良好的预测结果,准确率达90.3%。This article takes the second-hand housing price data of Chengdu in 2022 as the research object, and constructs a random forest model and XGBoost model to predict the second-hand housing price. Firstly, the dataset is cleaned and visualized to construct virtual variables. Then, a heat map is drawn and the entropy method is used for feature value screening to select important features for training the model. Subsequently, prediction systems based on random forest and XGBoost were developed using grid search techniques, and the accuracy of the models was measured using three key indicators: coefficient of determination, mean square error, and mean absolute error. After model comparison and result analysis, it was found that the optimized XGBoost model had good prediction results for second-hand housing prices, with an accuracy rate of 90.3%.展开更多
本文深入探讨了福州市财政收入的影响因素及其预测方法,首先运用描述性统计分析、Pearson相关性分析等方法,对影响福州市财政收入的关键因素进行了初步分析,揭示了各因素之间的相关性和变化趋势。其次,为减轻多重共线性对模型预测准确...本文深入探讨了福州市财政收入的影响因素及其预测方法,首先运用描述性统计分析、Pearson相关性分析等方法,对影响福州市财政收入的关键因素进行了初步分析,揭示了各因素之间的相关性和变化趋势。其次,为减轻多重共线性对模型预测准确性和稳定性的不利影响,本文采用了逐步回归法、岭回归法以及Lasso回归法来进行数据拟合。结果表明,Lasso回归表现出色,能够准确识别对财政收入影响最为显著的变量,这为政策制定提供了有力的实证依据。然后运用ARIMA模型对关键因素数值进行预测,得到2023年及2024年财政总收入预测值。最后,文章总结了研究的主要结论和政策建议,强调政府应加强对财政收入的预测和管理,制定科学合理的财政政策,以促进经济的可持续增长。This article deeply explores the influencing factors and prediction methods of Fuzhou’s fiscal revenue. Firstly, descriptive statistical analysis, Pearson correlation analysis and other methods are used to conduct a preliminary analysis of the key factors affecting Fuzhou’s fiscal revenue, revealing the correlation and changing trends between each factor. Secondly, in order to mitigate the adverse effects of multicollinearity on the accuracy and stability of model predictions, this paper used stepwise regression, ridge regression, and Lasso regression to fit the data. The results indicate that Lasso regression performs better and can accurately identify the variables that have the most significant impact on fiscal revenue, providing strong empirical evidence for policy-making. Then, the ARIMA model is used to predict the key factor values and obtain the predicted total fiscal revenue for 2023 and 2024. Finally, the article summarizes the main conclusions and policy recommendations of the research, emphasizing that the government should strengthen the prediction and management of fiscal revenue, formulate scientific and reasonable fiscal policies, and promote sustainable economic growth.展开更多
文摘本文以2022年成都的二手房房价数据为研究对象,构建随机森林模型和XGBoost模型来预测二手房价格。首先对数据集进行清洗并可视化处理,构建虚拟变量,接着绘制热力图并运用熵值法进行特征值筛选,选取重要的特征进行训练模型。接着,采用网格搜索技术分别开发了基于随机森林和XGBoost的预测模型,并利用决定系数、均方误差和平均绝对误差这三个关键指标来衡量模型的预测准确性,经过模型比较和结果分析,发现优化后的XGBoost模型对二手房房价有良好的预测结果,准确率达90.3%。This article takes the second-hand housing price data of Chengdu in 2022 as the research object, and constructs a random forest model and XGBoost model to predict the second-hand housing price. Firstly, the dataset is cleaned and visualized to construct virtual variables. Then, a heat map is drawn and the entropy method is used for feature value screening to select important features for training the model. Subsequently, prediction systems based on random forest and XGBoost were developed using grid search techniques, and the accuracy of the models was measured using three key indicators: coefficient of determination, mean square error, and mean absolute error. After model comparison and result analysis, it was found that the optimized XGBoost model had good prediction results for second-hand housing prices, with an accuracy rate of 90.3%.
文摘本文深入探讨了福州市财政收入的影响因素及其预测方法,首先运用描述性统计分析、Pearson相关性分析等方法,对影响福州市财政收入的关键因素进行了初步分析,揭示了各因素之间的相关性和变化趋势。其次,为减轻多重共线性对模型预测准确性和稳定性的不利影响,本文采用了逐步回归法、岭回归法以及Lasso回归法来进行数据拟合。结果表明,Lasso回归表现出色,能够准确识别对财政收入影响最为显著的变量,这为政策制定提供了有力的实证依据。然后运用ARIMA模型对关键因素数值进行预测,得到2023年及2024年财政总收入预测值。最后,文章总结了研究的主要结论和政策建议,强调政府应加强对财政收入的预测和管理,制定科学合理的财政政策,以促进经济的可持续增长。This article deeply explores the influencing factors and prediction methods of Fuzhou’s fiscal revenue. Firstly, descriptive statistical analysis, Pearson correlation analysis and other methods are used to conduct a preliminary analysis of the key factors affecting Fuzhou’s fiscal revenue, revealing the correlation and changing trends between each factor. Secondly, in order to mitigate the adverse effects of multicollinearity on the accuracy and stability of model predictions, this paper used stepwise regression, ridge regression, and Lasso regression to fit the data. The results indicate that Lasso regression performs better and can accurately identify the variables that have the most significant impact on fiscal revenue, providing strong empirical evidence for policy-making. Then, the ARIMA model is used to predict the key factor values and obtain the predicted total fiscal revenue for 2023 and 2024. Finally, the article summarizes the main conclusions and policy recommendations of the research, emphasizing that the government should strengthen the prediction and management of fiscal revenue, formulate scientific and reasonable fiscal policies, and promote sustainable economic growth.
