How to prepare and execute data

I have a machine learning program. Some code goes back to the 19902, inC. Including some linda code for a parralel execuyion (Linda) I can compile, execute etc but it's certainly not user friendly, nor programmer friendly

The data is "Nasty' The data is very skew and covers several orders of magnitude ,

. my trees usually 'prune' back to nothing.

Boosting gives invalid values of the 'targets. A REAL Mess.

The data I am testing seems to like pick up lots of values of a continuous variables and a 'two' step node, or multiple children selection , could be interesting

Apple appears to have a more modern version.

I can't seem to find out how to prepare the data and execute the code.

My retirement interest,