There are many ways to evaluate different strategies for solving different prediction tasks. In our last post, for example, we discussed calibration and descrimination, two measurements which assess the strength of a probabilistic prediciton. Measurements like accuracy, error, and recall among others are useful when considering whether random forest