prtClassAdaBoostFastAuc AdaBoost classifier (fast training)
 
    prtClassAdaBoostFastAuc is a version of prtClassAdaBoost that can
    be trained significantly more quickly than prtClassAdaBoost.
    prtClassAdaBoostFastAuc acheives this by assuming a linear
    classifier when picking the feature to be used in the weak learner.
    Unlike regular adaBoost, where the weak learner is trained and
    evaluated on each feature, in prtClassAdaBoostFastAuc, the feature
    is selected using simple ROC metrics, and this feature is used to
    train the weak learner for the current iteration.  This can be
    significantly faster than prtClassAdaBoost when the base learner is
    slow, or there are a very large number of features.
 
    a = prtClassAdaBoostFastAuc;
    a = a.train(prtDataGenBimodal);
    plot(a)