Rate This Document
Findability
Accuracy
Completeness
Readability

fault_train

Prototype

def fault_train(data_path, model_path, param)

Description

Fault training interface. It is used to train a new model.

Working Principle

Reads data and parameters to train a new model and dump the model file for prediction. In addition, this interface outputs the verification results of machine learning training metrics to a log file, including precision, recall (also called fault detection rate in the fault prediction field), FPR (also called false alarm rate in the fault prediction field), F1 score, and accuracy.

Figure 1 Metrics

Fault detection rate in the fault prediction field corresponds to recall in the machine learning field. False alarm rate in the fault prediction field corresponds to false positive rate in the machine learning field.

Parameters

Parameter

Description

Input/Output

data_path

Path to the data file to be trained

Input

model_path

Model file dump path

Input

param

Training parameters

Input