Hello,
I have a question. How are the examples divided when using X-Validation?
For example, when I use X-Validation with 4 iteriations on the data of 100 examples, does it mean that
- first 75 examples will be for the learning purpose and the next 25 for testing in the 1 iteriation, then 1-25 and 51-100 will be for learning and 26-50 for testing... etc. OR
- the set for learning will be each time 75 and for testing 25, but examples in sets are chosen randomly?
Thank you in advance
the second one is similar to Bootstrap validation. But in Bootstrap validation you can have examples twice in the training set.
~Martin