Mark A. Wolters

email me


Mark Wolters


Mark A. Wolters

Research faculty (青年研究人员)
Shanghai Center for Mathematical Sciences
22nd floor, Guanghua Tower East
Fudan University
220 Handan Road

Phone: 021-65643390
Fax: 021-65642190

Profiles: Google ScholarMicrosoft AcademicORCIDLinkedInResearchGate


Thank you for visiting. Here you will find links and information related to my research, as well as any other content of a technical nature that others might find interesting or useful.

My research interests are in the area of computational statistics. Much of my past work has involved the development of heuristic optimization algorithms for difficult constrained optimization problems in statistics and data analysis. More recently I have been working with probabilistic graphical models for binary data. I plan to keep building my capacity in statistical software development and high performance computing. I also have a background in engineering so I’m interested in applications to the physical and applied sciences. When I quit my engineering job to do grad studies in statistics, the term “data scientist” didn’t really exist. Whatever it means now, I’m pretty sure that’s what I wanted to be.

If you’re curious about anything you see here, feel free to email me.


R package scdensity has just been released on CRAN. It allows you to easily obtain kernel density estimates that satisfy certain shape requirements. For example you can get an estimate that must have only one mode, or, for extra smoothness, one that has two inflection points. You can see a little more on the code page.
My article Better Autologistic Regression has been published in Frontiers in Applied Mathematics and Statistics. I think it’s an important paper for anyone using autologistic regression for data analysis. If you happen to be one of those people, please check it out.
Added Interactive Demo: Autologistic Regression Models. The page contains a running Gibbs sampler that allows you to observe the behaviour of the autologistic model when you change parameters and model options. It is also my attempt to see what you can do in terms of interactive simulations with just javascript/html/css running in the browser (turns out you can do quite a bit).
I have created a LaTeX document class called ffslides, for creating presentations or posters. It is now available on CTAN. You can read a bit more about it on the code page.
The book chapter “Issues in the Identification of Smoke…” (coathored with Charmaine Dean) has just appeared. See the pubs page for details. Supplementary R code related to the chapter has also been posted today.

Added Sense and Separability in Logistic Regression with Biggish Data

Note: I’ve had intermittent issues with getting math to render on the page when accessed in China. Apologies to any visitors who might have the same problem. It’s likely a connectivity issue with due to the nature of internet access over here…