OMI特征提取算法及关于风险偏好的决策规则的研究

Research on OMI Feature Scoring and Statistical Decision Rule with Regard to Investor's Risk Preference

受到金融风暴的影响,投资者们的态度逐渐变得审慎,这使得金融机构迫切希望自己的投资产品能适销对路。通常情况下,风险与收益是成正比的,但是并不是每个人都能承受高风险,所以要抓住投资者的心理,向其推销适合的投资产品,需要先知道投资者的风险偏好。要做到这一点,必须在收集投资者的有关信息的基础上对投资者人群按风险偏好进行合理划分。本研究来自一份对投资者的调查问卷,设计OMI算法解决因子评分问题,使得属性因子的评分与风险偏好正相关;在此基础上用Kruskal-Wallis检验求出对风险偏好有显著影响的特征组合,从而产生最终的决策规则。

Because of the financial turmoil,investors have become increasingly cautious,which makes financial institutions eager to sell their financial products to the right customers.Under normal circumstances,profits is directly proportional to the risks,but not everyone can stand highrisk,so to identify the right customers,we need to know the risk preferences of investors.To do this,investors must be divided into different types according to their risk preference reflected by a collection of relevant information of investors.Based on an investor questionnaire,makes use of the OMI algorithm to solve the problem of feature scoring,making the score positively correlated with the risk preference;after that,Kruskal-Wallis test is used to find the feature combinations that significantly impact the investor's risk preference,resulting in the ultimate decision rules.