计算机科学与技术系

Department of Computer Science and Technology

Education background

Bachelor of Computer Science, Xi'an Jiaotong University, Xi'an, China, 1998;

Master of Computer Science, Tsinghua University, Beijing, China, 2001;

Ph.D. in Computer Science, Tsinghua University, Beijing, China, 2003.

Areas of Research Interests/ Research Projects

Evolutionary Computation

Distributed Intelligent Information Processing

Advanced Process Control for Manufacturing Integrated Circuits

National "No.02" Major Science & Technology Project: Detector Platform for PVD Equipment (2009-2010);

National "No.02" Major Science & Technology Project: Clustered Control Platform and Software (2009-2010);

National Natural Science Foundation of China: Modeling and Performance Evaluation Method of Cooperative Multi-Robot Systems using Agent-based Time Petri Net (2006-2007);

National Natural Science Foundation of China: Research on Learning and Cooperation Control of Cooperative Multi-Robot Systems using Probabilistic Model-Building Evolutionary Algorithm (2009-2011).

Research Status

My research team mainly focuses on evolutionary computation, intelligent information processing method, and integrating advanced process control technologies into circuit manufacturing equipments. In evolutionary computation, we work on the current estimation method by means of multi-variable distributed optimization algorithm, aiming at improving its convergence rate and the quality of optimal solutions. Our research on intelligent information processing utilizes machine learning algorithms to process Internet text messages and analyze the emotions hidden behind them. Research on advanced control technologies for integrated circuit manufacturing equipments focuses on developing clustered control software and platforms for IC manufacturing equipments.

We have proposed L1BOA, a multivariate Estimation of Distribution Algorithm (EDA), which uses L1-regularized regression to learn the structure of Bayesian networks. In L1BOA, structural learning is embedded within the process of parameter estimation, with the aim to build to a sparse but nearly optimized network structure. Compared with the classical Bayesian Optimization Algorithm (BOA), L1BOA is able to improve the efficiency of structure learning and automatically control the structure complexity.

Our achievements in the field of cluster control technologies for IC manufacturing equipments include etching machines for 200mm IC equipment and its control system. By 2007, we have successfully industrialized our technologies, applying them to the mainstream of 200mm and 300mm integrated circuit production lines. Our work in this field has received multiple science and technology progress awards.

Honors And Awards

National Science and Technology Progress Award, Second Class-Research, Development and Industrialization of 100nm High-Density Plasma Etching Machines (2009);

Science and Technology Progress Award by City of Beijing, First Class-Research, Development and Industrialization of 100nm High-Density Plasma Etching Machines (2007).

Academic Achievement

[1] CAI YunPeng, XU Hua, SUN XiaoMin, JIA PeiFa & LIU ZeHua, Duple-EDA and sample density balancing, Science In China Series F: Information Science, Sept. 2009, Vol.52, No.9, pp. 1640-1650, 2009.

[2] ZHAI Zhongwu, XU Hua, LI Jun & JIA Peifa, Feature Subsumption for Sentiment Classification in Multiple Languages,The 14th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD-2010), 21-24 June, 2010, Hyderabad India, (Accepted, to be published)

[3] Jiadong Yang, Hua Xu, Yunpeng Cai and Peifa Jia.  Effective Structure Learning for EDA via L1-Regularized Bayesian Networks. Proc 19th Intl. Conf. on Genetic and Evolutionary Computation (GECCO 2010), Portland, Oregon, 2010 (Accepted, to be published)

[4] Y. Wen, H. Xu, and J. D. Yang. A Heuristic-based Hybrid Genetic Algorithm for Hetergeneous Multiprocessor Scheduling. Proc. 19th Intl. Conf. on Genetic and Evolutionary Computation (GECCO 2010), Portland, Oregon, 2010 (Accepted, to be published)

[5] Hua Xu, "Petri Net: Theory and Applications "(Chapter 12: Timed Hierarchical Object-oriented Petri Net, I-Tech Education and Publishing, Vienna, Austria,2008,pp.253-280), ISNN:978-3-902613-12-7

[6] Hua Xu, "Recent Advances in Multi-robot Systems"(Chapter 13: A Novel Modeling Method for Cooperative Multi-robot Systems Using Fuzzy Timed Agent Based Petri Nets, I-Tech Education and Publishing, Vienna, Austria,2008,pp.249-262), ISNN:978-3-902613-24-0