金融大数据统计学习理论与方法及在互联网金融中的应用

面对大数据应用的快速发展、国家经济和金融安全所提出的迫切需求，我们面临着大数据分析方法瓶颈与挑战，需要发展大数据基础分析的理论方法和技术，同时应用这些理论方法研究大数据下的数据降维技术和算法，深入研究互联网金融风险管理、高频海量数据市场行为和管理决策等前沿问题。基于以上指导原则，本项目拟从以下五个方面对金融大数据统计推断理论及其应用等重大问题展开研究：（1）金融大数据“勘探”理论和算法；（2）金融大数据下高维降维技术的统计推断与算法；（3）金融大数据风险管理理论与方法研究；（4）高频大数据金融市场微观结构建模及其应用；（5）基于大数据的互联网金融计量建模及算法研究。这些研究内容相互联系与交叉，它们的核心是金融大数据计量建模和快速算法的提出。本项目拟通过这五个方面的研究，一方面，希望能在大数据金融计量理论和方法上进行创新，走向本领域学科研究的国际前沿。另一方面，本项目的顺利实施能够为我国金融

Currently, it is strongly needed to develop both fundamental theories and applied methods for analyzing big-data. Furthermore, efficient algorithms are also needed for applications in internet financial risk management, market behavior and decisions under high frequency trades, and so on. In this proposal, our research focuses on five topics, including (1)theories and algorithms for financial big-data mining (2)statistical inferences and computing algorithms based on dimension reduction of high-dimensional financial data (3)Risk management for financial big-data (4)Modeling micro structure of high-frequently trading financial market with its applications (5)Econometric models and algorithms for internet finances based on big-data technology. These topics are related each other, which emphasize on modeling and computation for financial big-data analysis. Our research will not only lead to some innovational theories and methods for big-data analysis, but also give feasible and solid suggestions on the security of China’s financial system.