基于交易数据的券商个人客户风险偏好识别

针对个人客户，基于客户的交易标的，应用主成分分析法、改进的K-means聚类算法为某证券公司的个人客户建立投资风险偏好模型。研究表明:该模型可为券商的客户细分提供精细化数据支撑，提升券商对客户风险偏好的认识，促进券商营销的精准度和成功率，为券商制订科学的营销策略提供决策支持。

Research on the risk preference recognition model of individual customers of brokerage firms based on transaction data

An investment risk preference model was established for individual customers of a brokerage firm, which was based on the customers’ transaction targets, the principal component analysis method and the modified K-means clustering algorithm. Results show that the model can provide detailed data support for brokerage firms to classify individual customers and improve their understanding of customers’ risk preferences, and promote their marketing accuracy and success rate, thus providing decision support for the firms to formulate scientific marketing strategies.