Dr. Tung-Lung Wu

Research Interests: 

Distribution theory of runs and patterns, high-dimensional hypothesis testing, control charts, scan statistics and cluster detection with applications to copy number variation (CNV) analysis.

Author Profile on Google Scholar
Author Profile on MathSciNet

Research Description:

My research develops probabilistic and statistical methodology for analyzing runs and patterns using the finite Markov chain imbedding (FMCI) approach. My work includes scan statistics for detecting localized clusters and anomalies, boundary crossing probabilities of stochastic processes, distributions of runs and patterns, and statistical process control, with applications to copy number variation (CNV) analysis, reliability, and quality engineering.

I am also interested in high-dimensional hypothesis testing for mean vectors and covariance matrices. My research investigates random projection methods for reducing dimensionality while preserving the essential statistical structure of the data, allowing classical inference procedures to be applied in high-dimensional settings. A major focus is the development of the new asymptotic theory required to establish the validity, power, and efficiency of these projection-based methods.

Software: FMCI Web App

A MATLAB Web App has been developed for computing the exact distributions of runs, patterns, scan statistics, and related stochastic quantities using the finite Markov chain imbedding (FMCI) approach. The web application provides an interactive interface for researchers and practitioners to perform exact probability calculations without manually constructing Markov chains. It is available at FMCI Web App. MATLAB installation NOT required.

Degree: Ph.D. in Statistics, 2012, Department of Statistics, University of Manitoba

Curriculum Vitae: Dr. Wu's CV