Chiefraven,
Anything you would like to know about WEKA can be found here:
http://www.cs.waikato.ac.nz/ml/weka/
I use the linux version and run everything via the command line. They have a GUI as well, but for the amount of data I process (many many gigs of data) the GUI is worthless.
One of the classifier trees is called J48. It will write rules for you in an IF=>THEN=>ELSE format. This will quickly let you see how you might adjust your parameters for various results. I use a combination of J48 for its simplicity and NaiveBayes. I've written my own wrappers to pre-process data, and to apply the classifiers against the data and save statistics to a directory. Then via a simple linux grep command I can see the % correct as well as other criteria and evaluate various models.
My other application is based on a genetic algorithm. There is a long story behind this one, and for the time being, suffice it to say that WEKA will address virtually all of your concerns and even some of the genetic algorithm stuff.
I would explore WEKA 100% before buying anything. The problem is that in order to use it you have to get your hands dirty and learn about parsing a text file, pre-processing your data, Java, etc.
Personally I think they are just a tool for confirmation and understanding. Some people have been able to get them to work to trade automatically. I haven't. Even those who do get them to work have documented that it is a constant battle dealing with degradation of performance and when to retrain.
I haven't quit my day job based on any of this so take everything with a grain of salt. Then again, only a fool would quit my day job, but that is another story.
Before you get too involved, allow me to make a comment. I've been trading on and off seriously for about 13 years and semi-seriously for an additional 10 years. In all this time, the only thing I have come across that can consistently get you in on the right side, make you money even if you make mistakes, and doesn't require much time or energy, is the Hershey method as espoused here by Spydertrader.
The only reason why I continue my data mining studies is to help me to understand what is going on and to hopefully create a profile of the ideal DU stock on the daily watchlist. It is not necessary to do this!
wash $60beer money! Thanks AU...
