Is it really this hard to find a dissertation?
Reasonably regular visitors to the site will know that I’m now in my fourth year of a mathematics degree at the University of Warwick. A good sized chunk of this year’s credit is given for a dissertation on a subject of our choosing; we can either choose to write about some kind of mathematics in action (the MIA project) or a research project, which involves finding a supervisor willing to sponsor you on a topic of your choosing.
As I’m hoping to continue further with my education, possibly going onto a PhD, I chose to do a research project. This post is basically an entry describing my particular woes finding an appropriate project, and it’s quite long; unless you’re interested, it’s best not to continue further!
My choice of project has certainly had its moments. Obviously, with my main interest being in the area of computational mathematics, I would like a project in this area. Beginning at the end of July, the search was on for a topic that I enjoyed and could write a decent-sized report on. Things turned pretty much from bad to worse – the guy that I had originally hoped to do a project with is on sabbatical this year, and so was obviously unavailable. I went through a series of supervisors over a two month period, none of whom were offering a project that was feasible for my current skill-set.
However, at the beginning of September, I thought that I did indeed have a suitable project lined up, to do with a relatively new (or at least untouched) area of computational mesh-free methods for PDEs (Partial Differential Equations). Rather, I would be working on some existing code and be part of an ongoing team developing some new and exciting stuff.
The idea is that one can model a complex domain, such as the interior of a cell or intake manifold of a car far more easily than creating a standard mesh and using some sort of finite element method. My job, in particular, was to make some kind of time-stepping routine which could be used to arbitrary order. As it turns out, there were a number of problems with this particular choice of project.
- The larger project was running under the guidance of four PhD students, all of whom had between one and two years experience with the existing code. The supervisor in the mathematics department had no direct role in the project, other than to organize group meetings and offer guidance and insights where necessary. Basically; I would have been working by myself, with the occasional help from one of the post-graduates. This was neither fair on them, nor me.
- The majority of the literature on the subject is divided up into about 6 papers which outline the broad methods. All of these papers were well beyond my ability levels in terms of understanding. The only book available was a rather complex book, involving things like Sobolov spaces, which I am yet to study; even this did not really touch upon the mesh-free methods aspect, rather some complex time-stepping routines.
All of this culminated in me going to see the R-Project co-ordinator at the end of last week. Basically, he said that for a 4th year student to collaborate on a project with post-graduates is pretty unfair, due to the large gap in knowledge and experience gap. This, combined with the potential lack of support from my supervisor lead me to contact the relevant people, and drop the project. It left me with somewhat of a bad taste in my mouth; not only did I have no project, but it was the end of week 3 and it meant that I’d effectively wasted a good two months.
However, a glimmer of hope came on the same day that I dropped the project. I got in touch with a professor from the Physics department, who also teaches a module on High Performance Computing. He was offering a title on Monte-Carlo methods, something that I really want to get involved in. It didn’t take me very long to realise this was the kind of thing that I was looking for in the first place really; writing a program from scratch to simulate a polymer in close proximity to an adsorbing surface and doing some basic analysis.
So, that’s where I currently stand. This weekend, I’m sitting down with a book and reading through some basic Monte-Carlo methods. It means that I once again have to deal with statistics, but since I can actually see some decent uses, I’m really quite looking forward to it.