基于Lasso-GA-SVR的区域物流业碳排放预测：以天津市为例

Carbon Emission Prediction of Regional Logistics Industry Based on Lasso-GA-SVR: Taking Tianjin City as an Example

摘要: 本文以天津市2000~2020年物流业的碳排放量数据作为研究对象，首先通过Lasso特征变量选择模型确定影响物流业碳排放的主要指标，然后将筛选出的指标值作为输入变量，建立基于遗传算法的支持向量机模型，以此构建Lasso-GA-SVR模型来预测天津市物流业的碳排放量。将该模型与Lasso-GS-SVR和PCA-GA-SVR模型进行对比，结果表明Lasso-GA-SVR模型具有更好的预测效果，据此利用该模型预测2021年天津市物流业碳排放量。

Abstract: This article takes the carbon emissions data of the logistics industry in Tianjin from 2000 to 2020 as the research object. Firstly, the Lasso feature variable selection model is used to determine the main indicators that affect the carbon emissions of the logistics industry. Then, the selected indi-cator values are used as input variables to establish a support vector machine model based on ge-netic algorithm, and a Lasso-GA-SVR model is constructed to predict the carbon emissions of the logistics industry in Tianjin. Comparing this model with the Lasso-GS-SVR and PCA-GA-SVR models, the results show that the Lasso-GA-SVR model has better predictive performance. Finally, the proposed Lasso-GA-SVR model is used to predict the carbon emissions of Tianjin’s logistics industry in 2021.