银行定期存款客户预测——基于Lasso-Forest方法 |
| |
引用本文: | 邓博. 银行定期存款客户预测——基于Lasso-Forest方法[J]. 科技和产业, 2023, 23(10): 151-157 |
| |
作者姓名: | 邓博 |
| |
作者单位: | 西安财经大学,西安 710000 |
| |
摘 要: | 从定性角度通过对数据进行描述性统计分析,给予银行营销策略建议。使用Lasso-Forest、Logistic、决策树及SVM模型分别进行建模分析,结合不平衡数据处理方法SMOTE算法,使用一系列评价指标进行模型效果评估,得到最优模型。所提出的Lasso-Forest组合模型具有比上述其他模型更精确的预测效果,识别准确率达到93%。使用提供的营销策略及Lasso-Forest模型可以有效降低银行成本,对客户类型进行精确识别并加以针对性营销,以此达到增加银行定期存款储备的目的。
|
关 键 词: | Lasso Forest 银行存款 分类预测 机器学习 |
Bank Term Deposit Customer Forecast: Based on the Lasso-Forest Methodology |
| |
Abstract: | From a qualitative point of view, through descriptive statistical analysis of the data, some suggestions were given for bank marketing strategies. Lasso-Forest, Logistic, Decision Tree and SVM models were used for modeling and analysis, combined with the unbalanced data processing method SMOTE algorithm, a series of evaluation indicators were used to evaluate the model effect and obtain the optimal model. The proposed Lasso-Forest combinatorial model has a more accurate prediction effect than the above other models, and the recognition accuracy reaches 93%. Using the provided marketing strategies and the Lasso-Forest model, it is possible to effectively reduce bank costs, accurately identify and target customer types, and achieve the purpose of increasing bank fixed deposit reserves. |
| |
Keywords: | lasso forest bank deposits categorical forecasting machine learning |
|
| 点击此处可从《科技和产业》浏览原始摘要信息 |
|
点击此处可从《科技和产业》下载全文 |
|