Hey everyone, you know what I've discovered. Is its easy to find information on Genetic Algorithms, and pretty easy to find an abundtant information on Neural Networks. But fuzzy logic to find any quality reports or comprehensive information seems to be very tough. It's just soo fuzzy :p So if anyone knows of good sites to visi that explains fuzzy logic and how it works. I would greatly greatly appreciate it.

Have you checked here?

The site itself is already about AI.

I didnt found any good sites on fuzzy logic.

But the book "Fuzzy Logic, Intelligence, Control and Information" by Yen&Langari is a good introduction I think. Check your local college library.
Good luck!

The book: "AI Game Development" by Alex J. Champandard is good too.

I checked my college library, and found out it really sucks. So did the cities library on this kind of stuff. So almost everything Ive done is from online as well as the book AI Techniques for game programming by mat buckland. Thanks for the help anyways.

As for generation5... the information isn't indepth enough that I need, and is very short. Again though thanks everyone for the help Im sure Ill find something.

I've been learning fuzzy logic recently, because it's become useful for one of my projects. The big thing I've noticed is that fuzzy logic is not the same thing that I expected it to be. It is not so much about probabilistic reasoning as it is about reasoning about inexact quantities. I can imagine some uses it might have within the context of game AI, but only on a rather advanced level, and it probably wouldn't be worth the added complexity. Nevertheless, if you want, I'll point you to the books I found helpful.

Man i hate this new look, well anyways. That would be great help if you could point me into direction with more advance fuzzy logic. Im not completely doing only Gaming AI as I wish to learn how it works in general and the many applications that it has.

Currently im reading a book by Champanard

Actually I find fuzzy logic to be marvellous for mob AI. It's basically a tolerance scale - you can make something very scary (scariness=1.0) or not very (scariness=0.2), give an entity bravery to counter scariness and use deterministic procedures on the result ((scariness-bravery)*-1) = bravado. With high bravado against you, a mob may attack, but then you can attach other fuzzy effectors to the bravado (like woundedness).

Really when you think about it, it's very useful indeed. Still boils down to an if...else... at the end of the day, but you can make the lead up to it very complicated, and seem more 'reactive'.

Quote: Original post by Sneftel
I've been learning fuzzy logic recently, because it's become useful for one of my projects. The big thing I've noticed is that fuzzy logic is not the same thing that I expected it to be. It is not so much about probabilistic reasoning as it is about reasoning about inexact quantities. I can imagine some uses it might have within the context of game AI, but only on a rather advanced level, and it probably wouldn't be worth the added complexity. Nevertheless, if you want, I'll point you to the books I found helpful.

That's a common misperception. Fuzzy logic has nothing to do with probabilities.
But it caused alot of controversy for years so it lost alot of popularity. It is a way of describing a continuous amount of membership something belongs to a set.

I find fuzzy logic very useful for games. Fuzzy state machines are very fun. Dissimilar to Probabilistic but kind of appears like it. Except that a FuSM fires in parallel all rules of membership, so it has some nice implications.

Anyways, look for FFLL (Free Fuzzy Logic Library). Nice little library in C++.

I used to be quite anti-FL, mostly because of my Bayesian background and training in probabilistic AI. However, I've since come to understand that FL definitely has its place (albeit a limited one, much as ANNs). It was the realisation that FL is equivalent to a B-spline representation of the domain (using triangular basis splines) that brought the most insight (for me) into its appropriate use.

At least as far as AI is concerned, one should not treat FL as a means of representing uncertainty in knowledge about a domain (for which probabilities are very appropriate) but rather as a means of representing vagueness about the actual attributes of the domain upon which knowledge is recorded.

Cheers,

Timkin