Help me decide what to learn!

January 4, 2009 10:56 PM

I have an empty space in the upcoming semester, my last as an undergrad (CS Major), and I am thinking about doing some sort of independant study/research, I just don't know what! I need ideas/inspiration! Help!

I have a professor who is willing to oversee something of this sort lined up already, but I am on my own as to coming up with what to do. I'm definitely interested in possibly branching out a little bit from what I have been doing for most of my college career (java, ruby/rails, more java) and working with an area where I am less familiar. Working with some sort of functional language, like Erlang or Haskell, sounds pretty cool. Also, the professor has a PS3 with linux installed on it that is currently acting as a 300 dollar paper weight in his office, so it could be fun to do something with that. I'm also interested in genetic algorithms, image processing, photography and the internet/large datasets.

So, if you've done something like this (independent study and/or research) before, how did you design the course for your self? How did you come up with the content of the course? What resources did you find useful? What would you do differently if you had to do it over again? What else should I think about?

Sorry if this question rambles a bit.

posted by pwicks to education (3 comments total)

Android programming might be interesting. As far as new programming languages you might want to look at Clojure or Scala. Haskell has been around for a long time, so it has some staying power. So has python. Clojure and Scala are JVM languages and there are jython and jaskell for the VM (pythong and haskell implementations). I've been playing around in scala.

What about doing 3d programming with Java Open GL?

If you're interested in Machine learning, I'd grab a copy of Peter Norvig's textbook. I had that for an A.I. class my last year and it covers most of the stuff you'd need. You'll want to focus on the statistical methods like decision trees, Bayesian networks, neural networks, support vector machines, etc. That stuff ties in with "internet/large datasets".
posted by delmoi at 11:14 PM on January 4

One tactic that I've used for designing (graduate level) independent study is to effectively design myself a course, syllabus and all, so that I have a set plan for each week's reading and a deadline for the final project.

You may find it helpful to zero in on a particular topic, say genetic algorithms, and then spend some time looking up different ways genetic algorithms are used. Most likely, one or two of these uses will jump out at you as being interesting, and then you work from there. (If none do, try another topic.)

And again, if you want to get the most out of this time, make sure you assign yourself readings for each week, don't just give yourself the final project to work on. It sounds like your professor wouldn't care what you did, but this will be your last semester to study whatever topics in the field you want...make the most of it!
posted by voltairemodern at 7:15 AM on January 5

Uh, your interest in "the internet" and big datasets along with your interest in image manipulation and photography could point you toward information visualization. You could check: infosthetics, flowingdata, neoformix, for some inspiration for things you could scrape from the internet and then put into a visual substrate (I'm thinking you could manipulate images in various dimensions to represent data from some data source). Here's an example project, InfoCanvas, that I helped out on as part of my CS research. Check the video for a good overview; you can show the boring papers to your professor.
posted by zpousman at 12:40 PM on January 5