Need Topic Suggestion for Project on Artificial Intelligence

What kind of things do they teach in 10th-11th grade math at your school?

Well, I may be in 10th grade but I'm in advanced classes so we're really doing 11th grade stuff. I already know about numbers, measurements, geometry, trignometry, algebra, probability. My classes will go through finite math, financial math, statistics, and precalculus followed by calculus.

The due date of this project is in May next year so we have like 8months to do this so by that time I think I'll know about calculus already. However, we need to pick a topic by Monday so I'm worried and am not sure what kind of stuff will be good for me. My teacher encouraged us to choose topics that we don't know much about yet but not TOO far above our level.

Also, it's not too late to repick a topic in another catergory so any suggestions on other math and science related catergory will be appreciated too.

Thanks.

Alright then, looks like you can do a lot of different topics. There is a list on Wikipedia at http://en.wikipedia.org/wiki/Artificial_intelligence#Mechanisms

You'll find a lot of statistics, probability, and measurements in Biology. If you want to look into computer encryption you'll find some interesting number theory behind it. I think you have a pretty wide range of options depending on how broadly you define "science".

Try speech recognition..
I work in the field of speech recognition. This is a pretty interesting part of the field of AI... and now it is actually being used commercially.

Statistics are a big part of the field... and enough of the math will be within your grasp to make it interesting (some you will have to skip).

I can point you to more information if this interests you.

Both speech recognition and computer encryption sounds fun but could you guys give me a specific problem in these fields and leave a little more information please? I need 3 topics. A brief explaination about how the math is applied behind it would save me a lot of trouble. Thank you again!

Also, a friend of mine suggested something like "Frame Problem of AI". I'm not really sure what that is or what kind of math is used behind it. Can someone explain it to me please?

5 classical problems in AI with mathematical models to help....

1. The Tower Of Hanoi Problem: You have 3 poles, Pole A (start), Pole B (intermediate) and pole C (End).Your job is to move N stacked discs placed in the increasing order of their diameter on pole A to Pole C using pole B temporarily.

Rules:
1. You should not disturb the disc's order on any pole. (i.e. a disc with smaller diameter cannot be placed below the longer one on any pole)
2. You can only move one disc at a time from one pole to another.
3. The number of attempts should be minimum.

input: No Of discs N
assume 1..N indicates the discs number in increasing order of their diameter

output:
Iterations will be the minimum number of attempts X

Attempt Number 1..X
Pole A Configuration: [Set of dics on pole A]
Pole B Configuration: [Set of dics on pole B]
Pole C Configuration: [Set of dics on pole C]

2. The N Queens Problem

Arrange N Queens on an N by N chessboard so that none will be able to cancel the other.

Input: N (number of queens)
Output: an N by N matrix dislayed showing single Q on each row such that none of the Q's cancels the other. there could be multiple solutions for the given N Queens on the N by N chess board

3. The Tic Tac Toe solver

One player will be human and other will be computer. Human player starts and writes an X into whichever box on a 3 by 3 matrix. then computer playes with an O and tries to either win or save one's own game.

Input: Program halts for human input where human writes the coordinates [x,y] wher he/she wants to put an X in the 3 by 3 matrix.
output: After the above human input, computer displays the matrix with its own O placed in a correct cell in order to either win or save itself.

4. The 8 puzzle problem (I think your friend's Frame Problem of AI is same as this ... it is probably the toughest but the most interesting of all)

More about this on wikipedia. I do have a VB6 solution to this If u want...

5. Several manifestations of the Shorted Path algorithm.

More about this on Wikipedia.

The "problem" in speech recognition is how to turn a series of sounds into the corresponding written text. There are several sub-problems... the most interesting is a matter of statistics. You have several possible options and the trick is to pick the most likely. Your brain dues thus atomatically so that even af I mike lats of mustakes you prutty mich know what I am treeing to say.

Think of how much a computer would have to understand to realize that "treeing" was really a error and that the word should be "trying" (not freeing or anything to do with oaks).

For a specific topic on this ... Bayes law would be a good place to start. There are plenty of good explanations of Bayes... including a wikipedia page.

An interesting related topic is "classification". This also is often done using Bayes law (which is why I thought of it.

Classification is the trick of looking at a group of data and deciding if it belongs in one classification or another.

An example is a Spam filter (very important lately) where you need to look at an email (or webpage) and simply by the words on it, decide whether it is spam or not. This is made more difficult because the people writing the spam don't want to be classified as spam.

Google uses similar techniques to figure out which web pages belong in the category you are searching for.

... and this is also used in reasearch to group data to find disease clusters etc.

This is a useful and very practical technology in AI.

As usual Wikipedia is a good place to start.

thanks guys! i'll bring up all these you guys suggested to my teacher and see what he thinks. it's likely i'll need some help later on too, i'll get back to you guys