Post-pruning using Error estimation

See more:

algorithm

I try to built a pruning method using the Error estimation but iam having difficulty to understand it. The link below is a link to an example of such method.

https://dl.dropboxusercontent.com/u/30465241/Capture.PNG[^]

Every node in the example has the bad and the good, which i dont understand what they are.

Also for the attribute membership says that f=5/14. Where did that come from?

If you have any ideas plz illuminate me.

Thanks a lot.