I can serve as dissertation advisor for a limited number of students at any given time. My research interests are rather broad over the field of computational and applied statistics. Applications in finance are of particular interest.
My students must be prepared for hard work. I do not believe in shortcuts; a student who wants to skimp on the normal procedures should seek a different advisor. My students are expected to know and to follow the rules and procedures set out in the University catalogue and supplemented by statements in the CDS Student Handbook. (There are professors who do not require that their students follow the University and Department requirements for the degree. Don't ask me who these professors are, but your fellow students also know who they are. If you're looking for this route, ask fellow students or recent graduates. In that case, do not ask me to serve on your committee.)
I like to work with students who have good backgrounds in mathematics. Mathematics is still the language of science. Although I do not do research in mathematical statistics, for the past few years, I have taught the math stat course sequence. I expect my students in statistics to take this sequence (and do well in it). For my students in computational finance, the course CSI 672 / STAT 652 is sufficient, although the sequence CSI 972 / STAT 972 -- CSI 973 / STAT 973 is preferred.
I believe both written and oral communication skills are important, and I will help students develop such skills. I expect these skills to be moderately well developed by the students' last years, and I will not write dissertations for students.
My students must have strong computing skills. In particular, they must be proficient in Unix (and maybe MS Windows also), in Fortran or C / C++, and in R / S-Plus or Matlab. (Fortran and R are the languages I use most often.) I also think all statistics students should learn SAS if they don't already know it. Just because.
I strongly prefer that my students use LaTeX (or plain TeX) for scientific writing. (Word or WordPerfect is good for writing memos or letters, however.) Of course, with TeX you need a good text editor -- and one with a good spell-checker. Macro or shortcut keys also help. (Actually, a word processor like Word can often be set up to do a pretty good job as a text editor.) On MS Windows machines, Crimson Editor is my current favorite editor. On Linux/Unix I still use emacs.
Several of my students have been part-time. While it took me a while to get used to the idea of a student pursuing a PhD degree part-time, I now realize that it is a realistic possibility. The student's full-time work often suggests areas for PhD research.
Informal associations among students outside of the classroom are important; not only for learning by working on statistics problems together, but also for forming networks that can be useful throughout one's career. This, of course, is another obstacle that faces the part-time student. Some of the CSI courses require group projects, and that helps to get students together outside of the classroom. At least a couple of times a year, I have a party at my home for the CSI CompStat and CompFin students and some local alumni.