Good Evening,
As I am playing with weka and learning I have a few questions I am not quite sure on the answer. I plug in my training and test data, run my algorithms and then I can get Weka to out put the probabilities. It looks something like this:
inst#, actual, predicted, error, probability distribution 1 1:win 1:win *0.514 0.486 2 2:loss 2:loss 0.464 *0.536
I assume, the numbers under the probability title is just the probability the classifier is correct?
What is "distribution"? It just looks like 1-probability.
What is the difference between confidence vs probability and how can I calculate it?
Thanks
[link][2 comments]