经济管理中复杂数据和复杂行为的分析方法及其应用

随着科学与技术进步，数据采集技术突飞猛进。管理者收集到海量数据的同时，数据类型和数据结构越来越复杂，数据中蕴含的信息繁杂多样。同时，加大了经济和金融环境的复杂性，直接影响经济与管理中个体行为和管理决策。本项目首先受社会失业调查中产生的偏差数据和中国家庭追踪调查中产生的缺失数据驱动，发展出一套创新的统计和计量分析方法。第二，针对经济与管理中出现的大数据及超高维数据，发展新的特征筛选方法和降维方法，剔除冗余信息，建立更加容易解释且预测能力更强的模型，并提出并行计算方法解决大数据带来的计算困难。第三，针对金融市场中投资者的复杂行为，我们借鉴行为金融中的理论构建计量模型和投资组合模型，系统地分析了投资者行为本身的复杂性以及投资者复杂行为与复杂交易环境的交互影响。最后，把新提出理论方法应用到应用经济、金融市场和医疗保险健康等领域。本项目的理论和实证研究在经济管理领域具有重要的理论价值和实际意义。

With the advancement of science and technology, data acquisition technology has advanced by leaps and bounds. As managers collect massive amounts of data, data types and data structures become more complex, and the information contained in the data becomes diverse and cumbersome. At the same time, it has increased the complexity of the economic and financial environment, directly affecting individual behavior and management decisions. This project is first motivated by the bias data generated from the social unemployment survey and the missing data generated from the Chinese household survey, and further develop a set of innovative statistical and quantitative analysis methods to deal with those two types of data. Secondly, for the big data and high-dimensional data recently arisen in the area of economics and management, this project proposes new feature screening methods and dimensionality reduction methods to eliminate redundant information and then establish models that are easier to interpret and have stronger predictive power. This project also proposes parallel computational method to tackle the computational difficulties brought by big data. Thirdly, aiming at the complex behavior of investors in financial markets, this project uses the theory of behavioral finance to build an econometric model and a portfolio model to analyze the complexity of investor behavior itself and the interaction between complex investor behavior and complex trading environment. Finally, the new theoretical models and methods are applied to the fields of applied economy, financial market and health insurance. The theoretical and empirical research of this project has important theoretical and practical significance in the field of economic and management.