Decision Tree (Classification/Regression) is a model to predict a sample’s classification or regression value. In structures of Decision Tree, leaf nodes give class labels and internal nodes represent certain attributes. Decision tree is a supervised learning algorithm, divided into two stages of training and inference. In the training stage, each partition is chosen greedily by selecting the best split from a set of possible splits, to maximize the information gain at a tree node. In the inference stage, a decision tree can clearly express the decision process, that is, after series of attributes testing, a sample finally reaches a leaf node.