Electrical Signal Processing

Dayu (David) Yang

QUALIFICATIONS

. Excellent knowledge in pattern classification, information retrieval,

image processing, anomaly detection, data mining, signal processing, and

statistical modeling.

. Great experience on support vector machine (SVM), neural networks, k-

mean, kNN, regression, decision trees, feature extraction, feature

selection, and decision fusion, etc.

. Great experience and understanding on image enhancement, restoration,

compression, segmentation and description.

. Strong problem solving skills. Solid Math background and excellent

knowledge of data structures and algorithms.

. Good understanding of the concepts of object-oriented software design.

Strong coding ability in C/C++ and Matlab. Great knowledge of MySQL.

Good Ubuntu experience.

. Highly self-motivated, inquisitive, creative and resourceful.

EDUCATION

Doctor of Philosophy, Electrical Engineering and Computer Science, 01/2006

- 05/2010

University of Tennessee, Knoxville

Master of Science, Electrical and Computer Engineering, 09/2001 - 12/2004

University of Tennessee, Knoxville

Bachelor of Science, Communications Engineering, 09/1992 - 07/1996

Chongqing University of Posts & Telecommunications, Chongqing, P. R. China

RESERACH AND WORKING EXPERIENCE

Research Scientist, 03/2010 - 06/2010

ExpRevenue Inc.

. Reviewed academic literature pertaining to content-based image retrieval

and data mining on text data. Participated in developing research

hypotheses. Applied image processing and pattern recognition methods to

build a commercial Chinese OCR system.

Graduate Research Assistant, 01/2006-12/2009

Advanced Imaging &Collaborative Information Processing (AICIP) Lab

Department of Electrical and Computer Engineering, University of Tennessee,

Knoxville

. Designed and implemented a highly efficient nonparametric quickest

detection algorithm which can be effectively applied to detect a wide

variety of anomalies or abrupt changes in real-time. Originally defined

a distance measure, Q-Q distance, for distributional change based on

quantile-quantile plot and provided supporting asymptotic analysis.

Designed and implemented a nonparametric decentralized quickest

detection procedure which, to the best of our knowledge, is the first

implementation of nonparametric decentralized quickest detection.

Developed C++ and Matlab code for a quickest detection system and

simulated the detection procedures.

. Designed and implemented a real-time algorithm for network intrusion

detection which includes feature extraction using independent component

analysis (ICA) and decision fusion to aggregate the classification

results from kNN and SVM classifiers. The KDD99 data was used and the

overall detection rate is 89.4%, which is comparable to the KDD99

winner's 92.7% detection rate but my algorithm is able to heavily reduce

the computational burden for classification tasks. Developed C++ code

for classifiers and data processing to realize the detection method.

. Designed and implemented a predictive model using auto-associative

kernel regression coupled with the statistical probability ratio test

and applied this model onto a simulated SCADA system to detect a variety

of common attacks or anomalies caused by human mistakes. This work

resulted in a patent applied by UTK.

. Researched on compressed sensing and implemented image compression and

reconstruction using compressed sensing technique. Also developed C++

code to implement image processing algorithms, including image

enhancement, restoration, compression, segmentation and description.

Graduate Research Assistant 01/2002-12/2004

Applied Visualization Center (AVC)

Department of Electrical and Computer Engineering, University of Tennessee,

Knoxville

. Implemented a 3D sound system using the Head Related Transfer Functions

and successfully simulated a virtual environment with multiple sound

sources. This is also my MS thesis work.

. Developed CheckViz, a 3D visual simulation software for airport security

training.

HONORS AND ACTIVITIES

. Excellent Graduate Student Award 2009

. President of Chinese Students and Scholars Association 2002-2004

PUBLICATIONS

1] Dayu Yang, Hairong Qi, "A Network Intrusion Detection Method using

Independent Component Analysis", IEEE 19th International Conference on

Pattern Recognition, Tampa, FL, Dec. 2008.

2] Dayu Yang, Alexander Usynin, and J. Wesley Hines, "Anomaly-Based

Intrusion Detection for SCADA Systems", 5th International Topical

Meeting on Nuclear Plant Instrumentation, Controls, and Human Machine

Interface Technology (NPIC&HMIT 2006), Albuquerque, NM, Nov. 2006

3] Dayu Yang, Alexander Usynin, and J. Wesley Hines, "Anomaly-Based

Intrusion Detection for SCADA Systems", IAEA Technical Meeting on Cyber

Security of NPP I&C and Information systems, Idaho Fall, ID, Oct. 2006

4] Dayu Yang, Hairong Qi. "An Effective Nonparametric Quickest Detection

Procedure based on Q-Q Distance", the 35th International Conference on

Acoustics, Speech, and Signal Processing, 2010

5] Dayu Yang, Hairong Qi. "An Effective Nonparametric Quickest Detection

Approach", IEEE Transactions on Signal Processing. Submitted for review.

6] Dayu Yang, Hairong Qi. "An Effective Decentralized Nonparametric

Quickest Detection Approach", IEEE 19th International Conference on

Pattern Recognition, Istanbul, Turkey, Dec. 2010.

243 Buena Vista Ave.

Sunnyvale, CA 94086

Tel: 408-***-**** (c)

Email: ********@*****.***

Contact this candidate