Table 5 The performance comparison of MB-LSTM with the baseline methods in KDEda and Cl.

From: MBLSTM is a contextual interaction refined method for time series prediction

Methods

Accuracy

Precision

Recall

F1-score

SVM + KDEda + Cl

0.701

0.896

0.752

0.818

GBM + KDEda + Cl

0.731

0.705

0.694

0.749

LSTM + KDEda + Cl

0.728

0.894

0.789

0.838

Bi-LSTM + KDEda + Cl

0.756

0.896

0.820

0.857

Mogrifier LSTM + KDEda + Cl

0.772

0.889

0.851

0.870

SAnD + KDEda + Cl

0.779

0.913

0.822

0.874

AdaCare + KDEda + Cl

0.78

0.918

0.838

0.869

MB-LSTM + KDEda + Cl

0.783

0.918

0.857

0.876

Back to article page