How good is my error in data mining

I'm trying to calculate how good are my measurements in machine learning! Let's say that I have five choices, and that error is 4,2, 0.002, 3, 6. Naturally, I will pick third one for the hit, but I would like to say following:

• I'm X% certain that hit is third pick

  I'm Y% certain that hit is first (last) pick

  Of course, X>>Y but I would like to know is there some kind of statistics metric for this.

I can always introduce my metric, but I'm interested is there some already known, so开发者_StackOverflow中文版 that I can avoid my explanation and just refer to particular metric???

The closest one is 'confidence interval', but that gives me the probability that my results are in some interval.

Thanks!

I think it depends on your machine learning algorithm, and your error measure.

Maybe you should ask on Statistical Analysis - Stack Exchange?