Skip to Main Content
Table 3

The specific details of two data processing methods and three machine learning algorithms

ModelOptimized model parametersModel inputsModel outputsApplication scenarioImplementation environment
DBSCAN EPS=0.066,
minpts=12 
All NWD data Abnormal data point Classification Python 
ADR-CI α%=97%,
β=2.17 
All NWD data Abnormal data point Classification Python 
ANN Number of layers=3,
Learning rate=0.02,
Activation=ReLU,
Batch size=80 
The 24-h NWD in the past few days The 24-h NWD of the target day Prediction Python 
RF Estimators=1,000;
min samples split=2,
min samples leaf=1 
The training data with 26 features The WD of the node Prediction Python 
XGBoost Estimators=2,000,
learning rate=0.05,
max depth=8 
The training data with 26 features The WD of the node Prediction Python 
ModelOptimized model parametersModel inputsModel outputsApplication scenarioImplementation environment
DBSCAN EPS=0.066,
minpts=12 
All NWD data Abnormal data point Classification Python 
ADR-CI α%=97%,
β=2.17 
All NWD data Abnormal data point Classification Python 
ANN Number of layers=3,
Learning rate=0.02,
Activation=ReLU,
Batch size=80 
The 24-h NWD in the past few days The 24-h NWD of the target day Prediction Python 
RF Estimators=1,000;
min samples split=2,
min samples leaf=1 
The training data with 26 features The WD of the node Prediction Python 
XGBoost Estimators=2,000,
learning rate=0.05,
max depth=8 
The training data with 26 features The WD of the node Prediction Python 
Close Modal

or Create an Account

Close Modal
Close Modal