The steps of verifying and evaluating the prediction model by using the verification data are as follows: 1. Using historical data to build prediction model: using historical data to build prediction, such as linear regression, logistic regression, decision tree or other classifiers. When establishing the model, the training data set should be used, and the data should be randomly divided into training data set and verification data set. 2. Using the model to predict the verification data: using the training data set to predict the verification data set, and comparing the prediction results with the actual results. 3. Evaluate the performance of the forecasting model: Different indicators can be used to evaluate the performance of the forecasting model, such as accuracy, recall, F 1 value, ROC curve and AUC. 4. Adjust model parameters: If the model performance is not good, it can be improved by adjusting model parameters. 5. Use the whole data set to verify the model: If the model performs well, you can use the whole data set to verify the model. This can increase the amount of data and further test the reliability of the model.