Add explanation of ‘random_tie_break’ to documentation

[fritshermans]

It’s not clear to me what the argument ‘random_tie_break’ exactly does. Could you please add a few lines to the documentation or examples about it?

[damienlancry]

in uncertainty sampling, we rank the data points by informativeness as measured by an acquisition function such as entropy or margin. In case there are several data points that are ranked equally, which ones should you choose? you can either take the ones returned by uncertainty.argsort()[:n] or you can first shuffle the data and then use argsort. that is what random tie break does here.