QBC stratified bootstrapping

[Wojtek-Rokicki]

I am writing my master thesis on imbalanced dataset impact on active learning methods. I have stumbled upon a problem in bootstrapping. When there is few minority class samples, sometimes models are bootstrapped only majority class samples and it results with: ValueError: The number of classes has to be greater than one; got 1 class

Would it be reasonable to implement stratified bootstrapping? I will make a pull request soon with the proposition.

[Wojtek-Rokicki]

#179