Was attempting to train a classification model on scikit-eLCS. In addition to not placing any rows in one of the two output classes, I noticed the training finished much faster than the prediction function.
That doesn't smell right to me, but I figured I should ask.