| 摘要 |
The present disclosure provides a classifier training method, including: acquiring a training sample set; determining a classification condition at a root node according to a preset classification condition feature, performing classification on training samples in the training sample set according to the classification condition at the root node, and acquiring a classification subset corresponding to a child node of the root node; using the child node of the root node as a current node, circularly implementing the step of determining a classification condition at the current node according to another preset classification condition feature, performing classification on training samples in a classification subset corresponding to the current node according to the classification condition at the current node, and acquiring a classification subset of a child node of the current node, till a leaf node; and determining a user identity classification result at the current node, and obtaining a decision tree classifier. The decision tree classifier obtained by using the classifier training method provided in the present disclosure has a good performance. The present disclosure further provides a classifier training apparatus, and an identity authentication method and system. |
