Lei Li

Assistant Professor

Ph. D., 1998, University of California at Berkeley

lilei@stat.fsu.edu

STATEMENT OF RESEARCH INTERESTS SELECTED PUBLICATIONS

 
 

STATEMENT OF RESEARCH INTERESTS
I was led to the field of statistics by its interdisplinary nature, and by my admiration for the great mathematicians Kolmogorov and Wiener.  After some years of study, I was able to appreciate the work of Mendel and Fisher.  The fascinating work of these giants demonstrates the endless power of human intelligence, and encourages us to continue the exploration.

My research interests are the "AAA" of statistical modeling:  applications, algorithms, and asymptotics.  In terms of applications, my work centers around two fields.  First, in statistical signal processing and time series analysis,  most of my work is on parametric models, and can be grouped into three modules: spectral estimation, spectral factorization, and deconvolution with nonnegative constraints.  Second, I have spent the last several years developing a model of the DNA sequencing  process, which plays a key role in current genetic research. I am also conducting some research in two other areas of genetics:  physical mapping and genetic mapping.  My other miscellaneous interests are scattered in the areas of  information theory, reliability, etc.

During collaborations with experts from other fields, I find that statistics is very necessary for current scientific research.   I enjoy those moments when I help others with my knowledge of statistics.
 
 


SELECTED PUBLICATIONS
  • Parametric Deconvolution of Positive Spike Trains,  Annals of Statistics 28, 1279-1301,  2000 (with Speed, T. P.). 
  • A Three-Color Labeling Approach for DNA Sequencing Using Energy Transfer Primers and Capillary Electrophoresis, Electrophoresis 19, 1403-1414, 1998 (with, Kheterpal, I., Speed, T. P. and Mathies, R. A.). 
  • Estimation of the Color Separation Matrix in Four-dye Fluorescence-based DNA  Sequencing,  Electrophoresis 20, 1433-1442, 1999 (with  Speed, T. P.).
  • Model Selection and Order Determination for Time Series by Information Between the Past and the Future, Journal of Time Series Analysis 17, No 1, 65-84,  1996 (with  Xie, Z.).
  • A Method to Factorize Spectral Densities of Multiple Moving Average Processes, Journal of Systems Sciences and Mathematics 7, No 2,  69-176, 1994.
This is an example of DNA sequencing data. (a) This segment was obtained by three color sequencing strategy. (b) We recovered the corresponding four color sequencing signal by a nonnegative regression and model selection procedure.
    a.
     
     
     

    b.
PREVIOUS PAGE                BROCHURE CONTENTS                      NEXT PAGE