Ask HN: Parallel Algorithms
2 comments
evolutionary optimization algorithms: particle swarm, ant colony, etc...
Nolite has a good suggestion... doing evolutionary computation in parallel can be fun. Another suggestion would be to look into graph algorithms and find something there. Graph algorithms, in general, aren't known for parallelizing very easily and there should be a fair amount of active research going on in that area. I imagine you could find something there that would have some value, and that might be fun.
I want to directly cut to the point: for a university project, I am required to think of an algorithm, implement it, make it run on a cluster in parallel and compare results, write a report, ...
I have a few ideas, but I already know those algorithms quite well and would like to use the chance to learn something new and exciting :)
The question is: which algorithms can you think of, that would be fun to choose?