By sampling your data, running the provided classifier on these samples in parallel on your own machine and letting your models vote on a prediction, we return much faster predictions than the regular machine learning algorithm and possibly even more accurate predictions.
