Service fault prediction method and device based on RF-XGBOOST

The invention discloses a service fault prediction method and device based on RF-XGBOOST, and the method comprises the steps: building a service monitoring index, and carrying out the preprocessing of collected index data, and forming a sample; dividing the sample data into a training set and a test set, respectively constructing an RF model and an XGBOOST model based on the training set, and constructing an RF-XGBOOST model through weight linear combination of the two models; the RF-XGBOOST model is evaluated on the basis of the test set; using an RF-XGBOOST model to monitor the service in real time, giving an alarm for a fault prediction result, and carrying out alarm processing; correcting and recording error prediction results through manual study and judgment; and real data in a period of time in the future is incorporated into a model sample. According to the method and the device, through continuous training of constructing an RF-XGBOOST model, advantages of machine learning and expert experience are c