用户登录 English
师资队伍栏目图
师资队伍大图
姓 名: 田博士
系 别: 信息管理与电子商务系
职 称: 助理教授
办 公 室: 工商管理学院办公楼227
办公电话:  
移动电话:  13808486941
E-mail: tianboshi@126.com

个人简介

    田博士,博士,现任湖南大学工商管理学院助理教授,硕士研究生导师。2014年10月毕业于香港理工大学,获得哲学博士学位。其主要研究方向为高维高频金融数据分析、投资组合与风险控制理论和最优化的理论与算法。已在Journal of Global Optimization、Optimization Methods and Software、Optimization Letters和Journal of Industrial and Management Optimization等管理科学与运筹方向的国际SCI期刊上发表论文6篇。

教育背景

2010.12---2014.10 香港理工大学  博士

2008.09---2010.09 湖南大学       硕士                     

2004.08---2008.06 湖南大学       本科                      

研究经历

2010.01---2010.08 香港理工大学 研究助理(Research Assistant)

2013.12---2014.10 香港理工大学 兼职研究助理(Part-time Research Assistant)

2010.12---2013.12 香港理工大学 助教(Teaching Assistant)

2014.10---2014.12 香港理工大学 副研究员(Research Associate)

 

招收金融工程,金融数据分析的研究生,有意愿者可以随时和我联系。

    热衷于金融工程和投资分析,对大数据的应用比较感兴趣;能独立思考问题,性格活泼,能吃苦耐劳的同学(文理兼收)。

 

讲授课程

统计学 计量经济学(双语和全英文) 多元统计分析和SPSS应用(双语和全英文) 运筹学 随机过程 大数据科学入门

投资理论与风险分析  金融时间序列分析 高频金融计量学

 

研究领域

高维高频金融数据分析 

投资组合与风险分析

最优化理论与应用

 

研究成果

近五年研究成果

 

1:Tian Bo-Shi(田博士)*,Li Dong-Hui and Yang Xiao-Qi, An unconstrained differentiable penalty method for implicit complementarity problems, Optimization Methods and Software, 2016,31(4):775-790,(SCI)影响因子:1.624, 运筹学与管理科学(JCR)

Abstract: In this paper, we introduce an unconstrained differentiable penalty method for solving implicit complementarity problems, which has an exponential convergence rate under the assumption of a uniform $\xi$-$P$-function. Instead of solving the unconstrained penalized equations directly, we consider a corresponding unconstrained optimization problem and apply the trust-region Gauss-Newton method to solve it. We prove that the local solution of the unconstrained optimization problem identifies that of the complementarity problems under monotone assumptions. We carry out numerical experiments on the test problems from MCPLIB, and show that the proposed method is efficient and robust.

 

2:Tian Bo-Shi(田博士)* and Yang Xiao-Qi, Smoothing power penalty method for nonlinear complementarity problems,Pacific Journal of Optimization,2016,12(2):461-484(SCI)影响因子:1.079,运筹学与管理科学(JCR) 

Abstract: In this paper, we introduce a new penalty method for solving nonlinear complementarity problems, which unifies the existing $\ell_1$-penalty method and the natural residual equation-based method. We establish the exponential convergence rate between a solution of the penalized equations and that of the complementarity problem under a uniform $\xi$-$P$-function and study a perturbed $b$-regularity condition. Two kinds of numerical algorithms with global and fast local convergence are designed by virtue of the proposed penalty method. Preliminary numerical experiments conducted on test problems from MCPLIB show that the proposed method is efficient and robust.

 

3:Tian, Bo-Shi(田博士)*, Yang, Xiao-Qi and Meng Kai-Wen, An interior-point$\ell_{\frac12}$-penalty method for the inequality constrained nonlinear optimization. Journal of Industrial and Management Optimization,2016,12(3):949-973, (SCI)影响因子:0.843,运筹学与管理科学(JCR) 

Abstract: In this paper, we study inequality constrained nonlinear programming problems by virtue of an $\ell_{\frac12}$-penalty function and a quadratic relaxation. Combining with an interior-point method, we propose an interior-point $\ell_{\frac12}$-penalty method. We introduce different kinds of constraint qualifications to establish the first-order necessary conditions for the quadratically relaxed problem. We apply the modified Newton method to a sequence of logarithmic barrier problems, and design some reliable algorithms. Moreover, we establish the global convergence results of the proposed method. We carry out numerical experiments on 266 inequality constrained optimization problems. Our numerical results show that the proposed method is competitive with some existing interior-point $\ell_1$-penalty methods in term of iteration numbers and better when comparing the values of the penalty parameter.

 

4:Tian, Bo-Shi(田博士)*, Hu, Yao-Hua and Yang Xiao-Qi, A box-constrained differentiable penalty method for nonlinear complementarity problems. Journal of Global Optimization,2015,62(4):729-747,(SCI) 影响因子:1.355,运筹学与管理科学(JCR)

Abstract: In this paper, we propose a box-constrained differentiable penalty method for nonlinear complementarity problems, which not only inherits the same convergence rate as the existing $\ell_\frac1p$-penalty method but also overcomes its disadvantage of non-Lipschitzianness. We introduce the concept of a uniform $\xi$-$P$-function with $\xi\in(1,2]$, and apply it to prove that the solution of box-constrained penalized equations converges to that of the original problem at an exponential order. Instead of solving the box-constrained penalized equations directly, we solve a corresponding differentiable least squares problem by using a trust-region Gauss-Newton method. Furthermore, we establish the connection between the local solution of the least squares problem and that of the original problem under mild conditions. We carry out the numerical experiments on the test problems from MCPLIB, and show that the proposed method is efficient and robust.

 

5:Li, Dong-Hui and Tian, Bo-Shi(田博士)*, n-step quadratic convergence of the MPRP method with a restart strategy, Journal of Computational and Applied Mathematics, 2011,235(17): 4978-4990. (通讯作者)(SCI)影响因子:1.112. 

Abstract: It is well-known that the PRP conjugate gradient method with exact line search is globally and linearly convergent. If a restart strategy is used, the convergence rate of the method can be an n-step superlinear/quadratic convergence. Recently, Zhang et al. [L. Zhang, W. Zhou, D.H. Li, A descent modified Polak–Ribière–Polyak conjugate gradient method and its global convergence, IMA J. Numer. Anal. 26 (2006) 629–640] developed a modified PRP (MPRP) method that is globally convergent if an inexact line search is used. In this paper, we investigate the convergence rate of the MPRP method with inexact line search. We first show that the MPRP method with Armijo line search or Wolfe line search is linearly
convergent. We then show that the MPRP method with a restart strategy still retains nstep superlinear/quadratic convergence if the initial steplength is appropriately chosen.
We also do some numerical experiments. The results show that the restart MPRP method does converge quadratically. Moreover, it is more efficient than the non-restart method.

 

6:Dai, Zhi-Feng and Tian, Bo-Shi(田博士), Global convergence of some modified PRP nonlinear conjugate gradient methods, Optimization Letters,2011, 5 (4):615--630. (SCI)影响因子:0.952运筹学与管理科学(JCR) 

Abstract: Recently, similar to Hager and Zhang (SIAM J Optim 16:170–192, 2005), Yu (Nonlinear self-scaling conjugate gradient methods for large-scale optimization problems. Thesis of Doctors Degree, Sun Yat-Sen University, 2007) and Yuan
(Optim Lett 3:11–21, 2009) proposed modified PRP conjugate gradient methods which generate sufficient descent directions without any line searches. In order to obtain the global convergence of their algorithms, they need the assumption that the stepsize is bounded away from zero. In this paper, we take a little modification to these methods such that the modified methods retain sufficient descent property. Without requirement of the positive lower bound of the stepsize, we prove that the proposed methods are globally convergent. Some numerical results are also reported.

 

 

科研项目

1:光滑化法方法求解互补问题以及在金融中的应用,湖南大学青年教师成长计划, 2015.01—2016.12,主持

2:均衡约束规划的新型松弛算法及其应用研究(11601142),国家自然科学基金,2017.01-2019.12, 主持

3:高维高频金融数据的实证研究(2016M602412),博士后基金面上项目,2016.12-2018.2, 主持

EMBA MBA MPAcc EDP
X

TOP