Assistant Computer Science

Guohua Hao

**** ***** *********** ****** *****: 541-***-****

Contact

School of EECS, Oregon State University E-mail: ****@****.***********.***

Information

Corvallis, Oregon 97331-5501, USA URL: http://web.engr.orst.edu/ haog/

Machine learning, structured especially sequential supervised learning; Conditional random fields

Research

(CRFs) with an emphasis on feature induction and scalability

Interests

Oregon State University, Corvallis, Oregon USA

Education

Ph.D., Computer Science (GPA 3.88) Sept. 2003 – Jun. 2009 (expected)

• Thesis Topic: Effective Training and Feature Induction in Sequential Supervised Learning

• Advisor: Professor Thomas G. Dietterich

University of Science and Technology of China (USTC), Anhui, P. R. China

B.E., Computer Science (GPA 3.84), Sept. 1998 – Jul. 2003

• Thesis Title: Automatic Text Location and Segmentation in Images and Video Frames

Proficient in C/C++, Java, Matlab, and Unix shell scripting

Skills

Extensive software development experience under Unix/Linux and Windows environments

Solid background in machine learning, mathematics, and statistics, familiar with machine learning

package Weka and statistical packages S-PLUS and R

Proficient in L TEX, Microsoft Office, and other common productivity packages for Windows and

A

Linux platforms

Experience with high performance computing clusters, MapReduce parallel computing paradigm

Fluent in both Chinese (native language) and English

Google Inc., Mountain View, California USA

Research

Experience

Software Engineering Intern Jul. 2007 – Oct. 2007

Worked on the “AskGoogle” query expansion project with Dr. Stefan Riezler.

• Built a query expansion system based on word sense disambiguation, where maximum

entropy model was implemented and trained from word-aligned query-snippet pairs.

• Designed and implemented the integration module which merges new query expansion

systems of AskGoogle project into the existing production query expansion system.

Oregon State University, Corvallis, Oregon USA

Graduate Research Assistant Feb. 2004 – Present

Worked on the project “Off-the-shelf Learning Algorithms for Structural Supervised Learning”.

• Used Conditional Random Fields (CRFs) trained by functional gradient tree boosting to

handle missing information in sequential supervised learning.

• Used Error Correcting Output Coding (ECOC) and non-independent training in CRFs to

deal with sequential supervised learning problems with large label sets.

Graduate Research Assistant Sept. 2003 – Jan. 2004

Worked on an intelligent desktop assistant project “Task Tracer”.

• Worked on the learning module of the system to track the work flow of users and predict

their future activities. (Team of 3)

1 of 2

National Laboratory of Pattern Recognition, Beijing, P. R. China

Visiting Student Sept. 2002 – Apr. 2003

• Designed and implemented algorithms for text location in video frames and static images.

(Advised by Dr. Hanqing Lu)

Department of Computer Science and Technology at USTC, Anhui, P. R. China

Research Assistant Mar. 2001 – Sept. 2002

• Worked on Content Based Image Retrieval (CBIR). Used Bayesian relevance feedback and

Independent Component Analysis for personalized image retrieval. (Team of 4)

• Explored several real time Linux systems, modified and designed parts of the modules for

a proposed embedded real time operating system. (Team of 3)

Oregon State University, Corvallis, Oregon USA

Teaching

Experience

Graduate Teaching Assistant Sept. 2007 – Jun. 2008

• CS162: Introduction to Computer Science II (Fall 2007, Winter 2008)

• CS331: Introduction to Artificial Intelligence (Spring 2008)

• CS534: Machine Learning (Spring 2008)

Thomas G. Dietterich, Guohua Hao, and Adam Ashenfelter. (2008). Gradient Tree Boosting for

Publications

Training Conditional Random Fields. Journal of Machine Learning Research, 9(Oct):2113–2139,

2008.

Guohua Hao and Alan Fern. (2007). Revisiting Output Coding for Sequential Supervised Learn-

ing. The Twentieth International Joint Conference on Artificial Intelligence (IJCAI-2007).

Jinlong Li, Guohua Hao, Weihong Wang, and Xufang Wang. (2004). Learning User Interest in

Image Retrieval. Mini-Micro Systems, Vol.25, No.7, P.1110-1112. (in Chinese)

TreeCRF: a C++ implementation of training conditional random fields (CRFs) with functional

Software

gradient tree boosting.

Package

Outstanding Student Scholarship at USTC, P. R. China. 1998, 2001, 2002

Honors and

Awards

Hua Wei Scholarship at USTC, P. R. China. 2000

A third prize in “Hua Wei Business Plan Contest” at USTC, P. R. China. 2000

Zhang Zongzhi Sci-Tech Scholarship at USTC, P. R. China. 1999

First-Grade prizes in National Mathematics Contest for high school students, P. R. China. 1996,

1997

Chinese Association of OSU

Miscellaneous

Activities

Treasurer 2005 – 2006

• Successfully held China Night 2006 (a party of 500 people from local community to celebrate

Chinese New Year) with other committee members.

Passionate, self-motivated, work effectively both independently and in a team environment

Personality

Available upon request

References

2 of 2

Contact this candidate