Episode description

AdaBoost is a canonical example of the class of AnyBoost algorithms that create ensembles of weak learners. We discuss how a complex problem like predicting restaurant failure (which is surely caused by different problems in different situations) might benefit from this technique.