So I'm taking an intro to AI class this semester and we get to choose our own semester project. I have a couple weeks to submit what I want to do, so I'm trying to generate ideas and see what's out there. So any great ideas? Part of my problem is I don't know what I will know how to do by the time I need this done, so I want something more compilcated than A*, but not so much I won't be able to finish. According to the syllabus we will cover things like searches (obviously), genetic algorithms, neural nets, and a few other things. My first idea was I wanted to do something with racing game AI, maybe cars that learn the best way to drive around a simple track. I know there are lots of resources on the net about this kind of thing so I'll probably check it out, but is that kind of thing really hard/advanced? Anyway, the project can either be a programming project or some kind of research paper, any ideas are appreciated!

I think your idea of making a self learning algorithm for a race track might be a good choice. It's simple and it would give you time to create a small graphical representation of it, if you'r felling lucky maybe even in 3d. =)

Perhaps not a perfectly shaded perfectly modeled car, in a great track, but something neat. =)

I asked my teacher about it and he pointed me towards RARS, which, coincidentally, I had just found after I emailed him. Looks really cool and would save me the work of everything except the AI itself, has anyone had experience with it? Is it hard to learn? It appears to be set up so you can do a learning AI if you want too, which would be cool.

I would suggest making a program that can play a decent game of Go.

While chess is more or less solved at this point, Go remains something where no good computer program has yet been written; possibly because of the elusive concept of "shape".

Quote: Original post by hplus0603
I would suggest making a program that can play a decent game of Go.

That's a good suggestion, however my friend is doing that one :). And while I don't think we can't do the same project (or similar), I'd like to do something different. We are allowed to work in pairs, however, so I'll keep that in mind.

I agree with the mod on learning the game Go.
The added benefit of working with a teammate would always help out enormously. It's always easier to get multiple ideas when working on a fairly busy project.

eheh

I'm downloading RARS =D

Quote: Original post by White Rabbit
I'm downloading RARS =D

Last news item on the first page there is dated 2003, but inside it says 2004. So...I guess they still have a season going :)

I don't think Go is a good idea for a first AI project. That'd be like taking an architecture course and trying to design the Gugenheim before you know what a brick is.

I second the RARs idea.

Goto your University library and read through some of the latest AI journals. Pay particular attention to the parts in articles which say what further work could be done and what could have been done better. There will be many obscure and abstract article topics, but you should be able to find a few concrete ones. That's how a lot of new research starts, by understanding the current progress in a particular field and trying to improve on it. I think that would make a very good impression on your professors as opposed to picking a random project or basing it on non-peer reviewed games material you find on the internet.