Hey, Ive been learnin to program for some time now. And now, I have to do a technical report so I've decided I want to get into Computer AI it gives me a good excuse to take the time to learn it while actually using it for class. So since this is the first stab I've done to learn Computer AI I was wondering if you have any recommended sites to visit, and books to read to learn. Please don't recommend google, Im not looking for a search engine. Thanks for your time, I'm looking forward to any responses that I get.

AI Junkie
AI Wisdom
Game AI

Should get you started :)

Jim.

Generation5

Thanks for the reply, they helped greatly I havn't really learnt much but I'm beginning to get some basic concepts that I need to look at down. Im creating myself a powerpoint presentation and am having troubles finding a good visual for a neural networking example. What would be really cool is if anyone has some code that they could post that I could make into an image. Or any other ideas for images to show how Neural Networking works, or where it is used.

Again I greatly appreciate all the help I've received so far.

This may not be exactly what you were looking for, but my neural net applets have a method that allows you to view the network. It doesn't really show how neural nets in general work, but it does look pretty cool :) .

http://24.179.70.51/Java/show/ai/network/
you can access the network view with enter

If you want where neural networking is used I have a general pattern finding applet as well:

http://24.179.70.51/Java/show/ai/network/test2.html


What type of code did you have in mind? As far as I know neural network code looks alot like every other type of code. Perhaps some code with the nn being used rather than code from the actual nn clases? Though that mostly consists of setting inputs calling a run function and taking outputs.

I'll throw some out anyways:





Hopefully something there will be useful...I'm willing to say a lot of stupid stuff if it increases my chances of saying something smart....

If you really want to get into AI and not just AI for game programming read the following books:

- Godel Escher Bach ( http://www.amazon.com/exec/obidos/tg/detail/-/0140289208/qid=1128283517/sr=8-3/ref=pd_bbs_3/104-4376459-9649531?v=glance&s=books&n=507846 )
- Russel & Norvig's: AI: A Modern Approach

GEB is heavy on the logic and stuff and is THE book that inspired many people to get into the field of AI, but might be a bit too heavy.

Russell & Norvig try to give a complete description of the field of AI it's not too complex but gives enough depth for the more serious reader.

Thanks, that GEB book may help me I do want to cover more then just GAME programming AI, though most of the time those are the only uses I can see it used. Though I have discovered some other, like robotics and other systems like expert systems. I've been reading through one of AI Techniques for Game Programming but for information and understanding I don't find it goes in very well. I don't like the way we have to do the report :p we have to do an outline of the major subjects and everything we are going to cover. But it doesn't make sense to me to do that before we do the research since we don't know what there is to cover!

I plan on covering genetic algorithms, fuzzy logic, and nueral networks. By the way thanks for the code samples and those other sites, though silly me forgot to come back and check I wound up getting some actual mathematics into the presentation w00t! Got a 95% or 98% can't quite remember but thats my best mark on a presentation soo far I was so excited!!!

Anyways, I'll go out and get that GEB book, I've got a month and a half to read it so hopefully It's not like 4000 pages lol otherwise Ill be spending alot of studying time haha.

Again thanks everyone for the help.

Heres the current outline that I am looking at, if anyone else could help me break it down better. I would greatly appreciate it.

1. Introduction
2. Definition of AI
3. Genetic Algorithms
3.1 Definition
3.2 How it works
3.3 Implementation
4. Neural Networks
4.1. Definition
4.2. How it works
4.3 Implementation
5. Fuzzy Login
5.1 Definition
5.2 How it works
5.3 Implementation
6. Conclusion

So far its pretty rough, if I break stuff down though it has to be in sets of 2 or more, I can't break something apart without having 2 subdivisions too it. Also anything I missed that I should make sure I cover?

You completely missed statistical A.I. topics such as bayesian classifiers, HMMs, etc. They can be quite useful - for example, many spam filters use bayesian classifiers. Decision trees (another statistical approach) and bayes nets have been used in a number of diagnostic applications, and have been generally successful. Here is an excellent resource on statistical algorithms - they are pretty much the lecture notes from the undergraduate and graduate A.I. classes that one of my professors made.

Edit: In fact, after looking at your list again, you hit some of the "popular" subjects people like to put into Game A.I. books and talk about on these forums, but you misses some more of the most useful algorithms. You should talk about path finding, from DFS to BFS to Dijkstra and finally A*. Take a look at the syllabus for our undergraduate A.I. course. It has a pretty good outline of what you should know if you are just starting out in A.I. (it includes genetic algorithms and neural networks, but in comparison those are small topics).

The only thing known for certain about AI is that as of yet it hasn't worked.

It's not meant to deter you from studying it, the field has contributed vast amounts of useful things to computer science, just not AI.