I am curious if anyone has a suggestion of an interesting algorithm in any segment of game theory. I need to do a research project for my algorithms class and I want to work on an interesting topic. The project is about 6 weeks long and does require implementation, so I don't want extremely complex or theoretical topics. Thanks for any suggestions.

Assuming you're not talking about the branch of mathematics called "game theory," I'm going to have to throw in with A*.

http://theory.stanford.edu/~amitp/GameProgramming/

Euler integration.

Quote:Original post by Woodchuck
It remember me a film with Russel Crow, a true story about a man that revolution the science of economy. If I remember well, this a game theory algorithm where agents must play together instead of against for better results.

Collaborative game theory?

Well, one interesting project I remember from my theory class was implementing RSA encryption. The basic idea and math behind it is reasonably simple, but if you need to make the project longer, you can always implement it for arbitrary key lengths. You'll need to code your own large integer data type for that, and implement basic arithmetic operations, like, you know, addition, multiplication, and for these to be reasonably fast, you will need to do a little bit of research into efficient algorithms of doing these. (Or, y'know, you could just find a bigint library on the net, but I assume you need something to code, so just throwing ideas out there.) Another thing you'll need to figure out is how to make up large prime numbers on the fly (no 'ideal' way, but some algorithms are more complex and precise than others, but of course, they run longer - one of the classic tradeoffs). So yeah, if making your own RSA system sounds exciting to you (it sure was to me), then you might want to look into that. ;)

Any way, I think it largely depends on what kind of stuff are you interested in. Pretty much anything in CS has math. So, figure out what kind of thing you would like to do - related to graphics, or AI, or security, or whatever, and once you have an idea there, it'll be far easier to find an interesting algorithm to work with, as I'd guess pretty much any algorithm that works with something you are interested with will be interesting. [smile]

Vovan

Thanks for all the ideas everyone. I decided to go with large-scale terrain rendering with adaptive LOD.