PhD Thesis - Nicolas Brodu's home page

I now completed all requirements for the PhD program in computer science of Concordia University (Montréal, Québec, Canada)!

Defence occured on June 22, 2007 (slides available), with the result of the thesis being accepted as submitted. This is a very rare outcome, most of the time the reviewers ask for minor modifications. One reviewer had actually never seen the like, and the department secretary had heard of it only once before. The examiners individually judged the thesis from "very good" to "excellent". Though I'm of course proud of the result, I keep in mind that my thesis is also a research report, that it can be improved and extended in several ways. I thus invite you to read my dissertation and give me any feedback you like, whether positive or negative, if possible constructive :)

My supervisor was Peter Grogono, a distinguished professor who is interested in particular in artificial life. I have thus joined him in september 2004 with the goal of defining my own subject in a related research domain. The external examiner was Wolfgang Fink, from the Visual and Autonomous Exploration Systems Research Laboratory at Caltech and the Jet Propulsion Laboratory. The other members of the committee were Gregory Butler and Joey Paquet from the Computer Science department, and Nawwaf Kharma from the Electrical and Computer Engineering department.

I worked on various aspects of emergent phenomena and complex systems. My approach on these controversial issues is a practical one (see the abstract below). I used Artificial Life and Artificial Intelligence as application domains for testing complex systems ideas predictively rather than descriptively. Some results can be found as separate projects or documents on this site, including:

Practical Investigations of Complex Systems is the title of my dissertation. Abstract:

What's currently called Complexity Science suffers from an unfortunate lack of consensus as to what is meant by these terms. A review of the common notions shows a field mined by controversies, with as many frameworks for the study of Complex Systems as there are authors who propose a generic one. This document is thus not an attempt to create yet another framework, but rather an application of the traditional scientific methodology to some Complex Systems in the domain of Computer Science. It is a demonstration that even for this field, the concrete application of predictive experiments set up to challenge the extent of the main notions proves fruitful. Moreover the tools and methods that are created along the way because they were necessary to carry on experiments represent by themselves an opportunity for making progress in the domain. This is precisely the case in the present Computer Science context in the form of new algorithms, that were successfully applied to the main experiments. Hence this work is both a call for a more classical and practical approach to Complexity, and a concrete application example for that call.

Since September 2007, I have joined the Bunraku research team as a post-doc researcher at the INRIA/IRISA institute (Rennes, France).