Computer Science Project Ideas?

[RandyWayne]

Whatever you do, trying to create a malevolent intelligence is generally NOT a good idea.

[jfrog]

Originally Posted by missourimary

jfrog, are you more into statistics, calculus, acounting, design, or geometry? What kind of math would you like your project to include?

not really accounting or geometry (though I've never taken the high level geometry course). but calculus, statistics, numerical methods, number theory, differential equations, proofs... i like all those.

Abstract Algebra and Linear Algebra are okay but not really my favorites.

[RandyWayne]

I tells ya..... a superior (to human) A.I. is the way to go! Just make sure you follow Asimovs Three Laws of Robotics.

[jfrog]

Originally Posted by RandyWayne

I tells ya..... a superior (to human) A.I. is the way to go! Just make sure you follow Asimovs Three Laws of Robotics.

lol, i was actually thinking i might do a project on artificial intelligence. A friend of mine showed me this and I was thinking I might can do it for my project.

http://battlecode.mit.edu/2010/info

[missourimary]

Wow, how badly I'm tempted right now...

Artificial intelligence, hmm? You could use [ you ] (remember the trick-but I'm NOT going to do it... I'm not, I'm not... I'm trying to be good...) as a model...

[CAD/JPY]

I would like an artificial intelligence program to judge human emotion in the stock market. Then I would like it to scalp small amounts of money, over and over again, never allowing a loss to take place... only profit. Each position being small enough that going to 0 wouldn't matter...

After that I would like to trade on the buy or sell side, and only when human emotion was at extreme sentiments.

I would like the AI to monitor the account, and adjust the size of trading positions as the account grew in value. Oh, yes... and as each order was triggered with a profit I would like the grid strategy to be re-opened as long as the human emotion was still at extreme sentiment.

I would also like the AI to have four seperate accounts. One for ultra-con, one con, one moderate, and one very aggressive. (Hint: will need backward induction) However, combined the accounts can never have more than 25% of equity at risk.... assuming hitting 0 is plausible.

I think that sums it up. Thanks! Email me when you have it ready.

[MrMasterMind]

A fashion design computer game, where if you put on the wrong garment you spontaneously combust?

[RandyWayne]

Originally Posted by jfrog

lol, i was actually thinking i might do a project on artificial intelligence. A friend of mine showed me this and I was thinking I might can do it for my project.

http://battlecode.mit.edu/2010/info

That looks like it would be a blast! Do you have to be a student to take part?

[Apocrypha]

Originally Posted by John Atkinson

Thats already been done. Everyone already knows its 42.

+1 to John.


So long and thanks for all the fish.