哈尔滨工程大学校对论文模板教材.docVIP

学 号2011113118 密 级 公开 哈尔滨工程大学论文 院（系）名 称：专 业 名 称：学 生 姓 名：指 导 教 师：  学 号2011113118 密 级 公开 分布鲁棒投资组合优化模型研究 学生姓名：王彪 所在学院：理学院 所在专业：信息与计算科学 指导教师：许丽艳 职称：讲师 所在单位：哈尔滨工程大学 论文提交日期：2015年6月 论文答辩日期：2015年6月 学位授予单位：哈尔滨工程大学 摘要min-min可以联合运算的事实，将模型转化为线性半定规划问题。并且根据Erick Delage和Yinyu Ye在文献[1] 的一些工作,给出了对以矩阵不等式表示的均值和协方差矩阵置信域的估计。利用来自上海证券交易所的六种不同行业的资产的股票的历史数据进行数值实验，讨论了模型的实用性。 关键词：投资组合优化；分布鲁棒优化；线性半定规划；锥对偶理论；置信域 ABSTRACT Portfolio optimization problem, which is one of the important problems in financial mathematics, is the process of choosing the proportions of various assets to be held in a portfolio, in a manner that trades off the expected returns and the risk of potential portfolios. Due to the uncertainty of the market, the portfolio optimization problem involves uncertain parameters usually. Thus, it has fairly theoretical and practical significance to study the portfolio optimization problem with uncertain parameters. In this paper, we focus on the portfolio optimization problem with uncertain parameters. Assuming that the parameters are multi-dimensional random variables with only partial information of the distribution of the parameters available, i.e., the bound for the mean and the covariance of the parameters, this paper attempt to find the best of the worst-case expected returns among all possible distributions, thus this paper establish the distributionally robust model for portfolios optimization. Because the inner problem of the model is a semi-infinite programming with the distribution (measure) as its variable, it is quite difficult to solve the model. This paper use the cone duality theory to convert the inner problem to its dual problem equivalently, which can be written equivalently as a linear semi-definite programming, and also use the fact that the