Hi, I am new to AI and I an currently looking for good tutorials. I know little about it and the math is difficult for me to grasp so I just wanted to know where I should begin? Do I just look for tutorials on machine learning? My main interest is in genetic programming or grammatical evolution for use in financial forecasting or trading system development. I came across a site a while back: http://www.tradingsystemlab.com/home.aspx and became interested in this type of application. That particular software is based on Discipulus and is very expensive ($60k). I don't have that kind of money so I would like to replicate something similar. Software that writes software is intriguing to me, so if anyone has some information on good tutorials, please let me know. Thank you.

This isn't really an appropriate place for your question, unfortunately. Game AI has little to do with financial AI.

There really aren't many AI forums out there, especially those with emphasis on financial AI. The few that there are rarely get posted on. But if you could point me in the right direction, I would be grateful.

People have been using the open-source software FANN (Fast Artificial Neural Network) library for stock forecasting. In fact, there is a recent thread on the subject in their forums: http://leenissen.dk/fann/forum/viewtopic.php?t=152.

I learnt the Neural network and genetic programming by a book, <AI for Game Developers>, published by O'Reilly. and Mat Buckland's AI Tutorials.

Hi. I've also bought that book, and tought the chapter on neural networks looked interresting, but it seemed to me like it was a bit too little info, for me anyway, as i'm all new to neural networks. My programming skills are adequate, and the math bit is also in place, but I was missing a little more info on the mechanics and techniques. Are there any other books out there that you would recommend reading? 0o

Yep I'd say that Matt Buckland's book AI Techniques for Game Programming is a great place to make a start on Neural Net and GA's. He has made a lunar lander app that uses GA's to evolve a solution which is pretty cool, has the source code available too.

I've made a Genetic Algorithm application myself that is similar to Buckland's and lands a space craft. It's a little prettier though and the UI is a little better :) Can send it on if you want it.

If you're confident in your programming skills and could turn theory into a program, I used a book called Machine Learning in university. It dives into what a neural network is and the portions of the solution space they can solve. It's also loaded with other forms of machine learning, and is a palatable read. 30 customer reviews at Amazon give it an average rating of 23, and since customer reviews are the source of all truth, you can be guaranteed that it is a good book. I really did enjoy it, and years later, I still read through it to remember things.

Again though, it's gets into the theory, and I don't remember there being any code. The algorithms are very clearly written out in pseudo-code. You'll come away with an understanding of what neural networks are/what they can do as opposed to some code you may have copy/pasted.

*Edit* - GAs are also explored in the book.