using radial basic function for prediction

I have a large data set(1829559*4) relating to "run to failure" of a device. three columns are my input an last column is my target value which I want to predict it (for unseen data). I am trying to use newrb for prediction however it is completely vague for me. I try to search the tutorial based on Mr.Greg contribution but I can not find it. please somebody help me to find a good tutorial. Furthermore, is there any suggestion of other techniques suitable for my problem? it is worth to say for each column in my data set the value of corresponding row is very very close to the previous row and I think this issue make the training unreliable.

Answer 1

[Bernhard Suhm]

You can find information about our implementation of radial basis networks in the documentation . There are many tutorials available on the internet, just search for radial basis networks.

[Greg Heath]

basi"S"

NOT

basi"C"

My tutorials are in the NEWSREADER, not ANSWERS.

comp.soft-sys.matlab

Greg