[Feature request] Adding MCC as a ThOpt_metrics's option

[lucazav]

According to this this article, the Cohen's k may lead to incorrect results:

"Despite its popularity, in the learning context there are a number of issues causing the kappa measure to produce unreliable results (for instance, its high sensitivity to the distribution of the marginal totals [73–75])

So, it may be useful to add the Matthews correlation coefficient as an option for the ThOpt_metrics parameter in the optimize_threshold_from_predictions() function.

[greglandrum]

Hi @lucazav, I agree that this would be a useful addition to the software. The next time I'm working on it, I'll try to remember to add it. Or, of course, pull requests are always welcome if you want to do it yourself. :-)