基于机器学习的客户资产配置决策偏好预测研究

资产配置决策偏好的准确预测有助于提升客户决策效率和满意度,降低金融机构的人力成本。通过构建基于机器学习客户资产配置决策偏好模型,设定输入变量为人口统计变量、家庭经济情况、个性心理特征和风险态度等四大类,输出变量为客户资产配置决策偏好选择,运用决策树和支持向量机等机器学习算法预测客户的资产配置决策偏好,并与传统预测方法进行比较。研究结果表明,机器学习算法能够在一定程度上预测客户的资产配置决策偏好,且其性能相较于传统预测方法更为有效。

Research on the Prediction of Decision Preference of Customers’Asset Allocation Based on Machine Learning

Accurately identifying decision preferences of personal asset allocation can help improve customers’decision-making efficiency and satisfaction,and reduce the labor costs of financial institutions.We construct the model of preference selection of customer’s asset allocation decisions based on machine learning.The input variables are demographics,family economics,personality psychological characteristics,and risk attitudes,and the model’s output variable is the preferred selection of customer’s asset allocation decision.The machine learning algorithms of decision trees and support vector machines are used to identify the customer’s asset allocation decision preferences.In addition,we also compare them with the traditional identification methods.The results show that the machine learning algorithm can identify customers’asset allocation preferences,and the performance is more effective than the traditional recognition methods.