Daniel Heinz

About Me

I received my M.S. and Ph.D. from Carnegie Mellon University. As an undergraduate, I attended Haverford College near Philadelphia, PA. At Haverford, I majored in Psychology, with a Math minor. Choosing a major was not easy, due to my diverse interests. In addition to Psychology and Math, I took several classes in Chemistry, Biology, and Foreign Languages.

Before beginning my graduate work, I was a member for Habitat for Humanity of Iredell County. I have had the distinct pleasure of taking part in almost every stage of construction from laying out a foundation to nailing in that last piece of baseboard in the bedroom closet.

I ran competitively for Haverford College, who won the NCAA 2010 national championship in cross-country. I still enjoy running for physical and psychological well-being. I am a fan of games of all shapes and sizes, from sports and card games to strategy, puzzle, and word games. As I suspect is true of many of my contemporaries, it was video games that first introduced me to the joys (and frustrations :) of programming.

Reading is another one of my hobbies. I recently finished Les Miserables. If you enjoy the musical you will probably enjoy this book immensely. You can read or skip the political diatribes according to taste. I also recommend my two favorite books: Catch-22 and The Count of Monte Christo.


I am currently an assistant professor at Loyola University of Maryland where my chief duty is decreasing the mean age of the statistics faculty. The median, of course, has not changed as much, which is a useful example for teaching about robust statistics.

I use visual aids in a lot of my classes to help students understand important concepts. Here are some examples that I created using R.


My current research involves computational statistics and Bayesian non-parameteric processes. As part of my dissertation research, I developed theory of a graphical version of the famous Dirichlet Process. Graphical in this sense refers to a set of independence relationships among observed variables. I apply this theory to make inference about Bayesian mixture models under conditional independence constraints. By comparing the marginal likelihood of a set of data under various graphical models, I determine the relationship between variables from a mixture of distributions. I have written code for these two applications, which is available here. In addition to working with the graphical Dirichlet Process, I am interested in expanding the theory to other extensions and applications, including the Hierarchical Dirichlet Process of Teh, et al., and Pitman-Yor Processes. Graphical versions of the Beta Process are also particularly exciting.

It may be a cliché, but I am generally eager to learn about almost any topic. I have worked at various times as a teaching assistant in a chemistry lab, a research assistant in a cognitive psychology lab, and as a proof-reader and solutions-writer for a topology textbook. I hope to become more involved in interdisciplinary research in the near future, especially regarding cognition and linguistics.

Some Useful Links