Brief Biography

Lingfei Wu is a 6th year Ph.D. candidate in the computer science department at College of William and Mary, advised by Dr. Andreas Stathopoulos.  Before coming to College of William and Mary, he received his M.S. degree in Automation from University of Science and Technology of China (USTC), following his B.E. degree in Electrical Engineering from Auhui University, both in Hefei, China.

Lingfei has been doing his internship at IBM T. J. Watson Research Center since May 2015. He is working with Dr. Jie Chen and Dr. Richard Lawrence to develop an efficient one-vs-one kernel ridge regression method via randomized features to scale up large-scale kernel machines for speech recognition. The proposed method can achieve much better classification accuracy and comparable recognition error compared to the best reported deep learning results, while requiring less memory and computational costs. In addition, he is developing a high-performance machine learning software for large-scale speech recognition tasks by leveraging hybrid MPI/OpenMP and MPI/GPU.

Lingfei had also been working with Dr. Kesheng (John) Wu since May 2014 as a computing sciences summer student in the Computational Research Division at Lawrence Berkeley National Laboratory. During his summer internship, he proposed a high-performance outlier detection method for finding blob-filaments in real fusion experiments and numerical simulations. In addition, he developed a real-time blob detection approach through hybrid MPI/OpenMP parallelization and this is the first research work to complete blob detection in 2 or 3 milliseconds.

Note: this website is not maintained anymore. Welcome to visit Lingfei's new personal webpage!

CURRICULUM VITAE [PDF]
    

Research Interests

Lingfei’s research interests are in the areas of numerical linear algebra, scientific computing, mathematical software, large-scale machine learning,  big data analytics, and high performance computing, specifically shown as follows:

• Numerical linear algebra, scientific computing and mathematical software

• High-performance eigensolver and SVD solver software for very large-scale problems
  
• Extreme singular value problem, especially for smallest singular triplets
• Extreme and interior eigenvalue problems
• Statistical learning for estimating trace of implicit matrix

• Large-scale machine learning and big data analytics
  

• High-performance machine learning algorithms and software
  
• Big data mining algorithms and software
  

• Parallel and high performance computing

• MPI, MPI/OpenMP for large-scale SVD problems
  
• Hybrid MPI/OpenMP and MPI/GPU for speech recognition
  
• Hybrid MPI/OpenMP for real-time blob detection in fusion plasma

News

06/2016: Teddy has successfully passed his dissertation defense ! New !

06/2016: Teddy was selected for SIAM Student Travel Award in SIAM AN 2016. New !

05/2016: Our paper "Revisiting Random Binning Feature: Fast Convergence and Strong Parallelizability" was accepted to KDD 2016. New !

03/2016: Our abstract paper "Preconditioned Thick-Restart Lanczos Method with Subspace Optimization for Symmetric Eigenvalue Problems" was presented at 14th Copper Mountain Conference on Iterative Methods.

01/2016: Teddy was invited to give a talk at the minisymposium entitled "Innovative Algorithms For Large-Scale Eigenvalue Solutions" for SIAM Annual meeting. 

01/2016: Our abstract paper "Preconditioned Thick-Restart Lanczos Method with Subspace Optimization for Symmetric Eigenvalue Problems" was accepted to 14th Copper Mountain Conference on Iterative Methods. 

01/2016: Our paper "Efficient One-Vs-One Kernel Ridge Regression for Speech Recognition" was accepted to ICASSP 2016. 

12/2015: Teddy has successfully passed his dissertation predefense ! 

11/2015: Our paper "High-Performance Algorithms For Large-Scale Singular Value Problems and Big Data Applications" was presented at SC15. 

10/2015: Our abstract paper "Improving Thick-Restarting Lanczos Method by Subspace Optimization For Large Sparse Eigenvalue Problems" was presented at SIAM LA 2015. 

09/2015: Teddy's internship at IBM has been extended to December 2015 due to his excellent performance !

09/2015: Teddy was selected for ACM’s SRC Travel Award for SC15.

09/2015: Our paper "High-Performance Algorithms For Large-Scale Singular Value Problems and Big Data Applications" was accepted to ACM/IEEE SC15 Doctoral Showcase.

09/2015: Our ACM student research competition poster "A High-Performance Preconditioned SVD Solver for Accurately Computing Large-Scale Singular Value Problems in PRIMME" was accepted to ACM/IEEE SC15.

08/2015: Teddy's summer intern work entitled "Scale Up Large-Scale Kernel Machines for Speech Recognition" was presented in the Summer Intern Seminar 2015 at IBM T. J. Watson Research Center.

06/2015: Teddy was selected for SIAM Student Travel Award. 

06/2015: Our abstract paper "Improving Thick-Restarting Lanczos Method by Subspace Optimization For Large Sparse Eigenvalue Problems" was accepted to SIAM ALA 2015. 

05/2015: Teddy is starting his summer internship at IBM T.J. Watson Research Center. 

04/2015: Our journal paper "A Preconditioned Hybrid SVD Method for Computing Accurately Singular Triplets of Large Matrices" was accepted to SIAM Journal on Scientific Computing. 

03/2015: Our abstract paper "An Implementation and Analysis of the Refined Projection method For (Jacobi-) Davidson Type Methods" was presented at SIAM CSE 2015. 

01/2015: Teddy was selected to receive a SIAM Child Care Grant for SIAM CSE15 :) 

01/2015: Teddy has successfully passed his dissertation proposal ! 

12/2014: Teddy was selected for Student Activities Conference Fund Award 2015, College of William & Mary.

11/2014: Our paper "High-Performance Outlier Detection Algorithm for Finding Blob-Filaments in Plasma" was presented at ACM/IEEE SC14. 

11/2014: Teddy was selected for Arts & Sciences OGSR/GSA Conference Funds 2014, College of William & Mary.

10/2014: Our paper entitled "High-Performance Outlier Detection Algorithm for Finding Blob-Filaments in Plasma" was accepted to Proc. of 5rd International Workshop on Big Data Analytics: Challenges and Opportunities (BDAC-14), held in conjunction with ACM/IEEE SC14.

09/2014: Our abstract paper "An Implementation and Analysis of the Refined Projection method For (Jacobi-) Davidson Type Methods" was accepted to SIAM CSE 2015.

09/2014: Teddy was selected for SIAM Student Travel Award.

09/2014: Teddy was selected for ACM’s SRC Travel Award.

09/2014: Our ACM student research competition poster was accepted to ACM/IEEE SC14.

07/2014: Our abstract paper "Accurate Computation of Smallest Singular Values Using the PRIMME Eigensolver" was presented at PMAA 2014.

05/2014: Teddy is starting his summer internship at Lawrence Berkeley National Laboratory.

04/2014: Our paper "Enhancing the PRIMME Eigensolver for Computing Accurately Singular Triplets of Large Matrices" was presented at 13th Copper Mountain Conference on Iterative Methods.

02/2014: Teddy was selected for Conference Travel Grant Award 2014, College of William and Mary.

02/2014: Teddy was selected for Student Travel Grant Award at Copper Mountain Conference 2014.

12/2013: Our abstract paper "Fitting techniques for estimating the trace of the inverse of a matrix" was presented  at ERCIM 2013.

06/2013: Our abstract paper "Using ILU(0) to estimate the diagonal of the inverse of a matrix" was presented at ILAS 2013.

03/2012: Our abstract paper "A MATLAB Interface for PRIMME for Solving Eigenvalue and Singular Value Problems" was presented at SIAM CSE 2013.