Experiment: From the above study, we can know that the core of ID3 should be in the decision tree at each node, the corresponding information gain criteria to select characteristics, recursively build the tree. The specific method is: starting with the root node (root node), calculating the information gain of all possible features, selecting the feature with the largest information gain as the node feature, establishing the child node by the different values of the feature, then calling the above method on the sub-junction recursively, building the decision tree, until the information gain of all features is small or no features to be selected, and finally a decision tree is obtained.<br>The training set is as follows:
正在翻译中..