In this book two methods are explored that are used for classification tree to handle large amount of data. In the first method to mine large data sets, a single training data set is be broken into n subsets. A classification tree will be learned from each of these n subsets in parallel. Rules are then be generated from the classification trees. These rules are combined into one rule set. Then the final set of rules are used to classify future unseen data. In the second method, post process the mined information from classification tree in order to extract actions to change the status of classified groups (e.g. customers) from an undesired status (such as attritors) to a desired one (such as loyal).
