Table 12 Efficiency of deep learning models on Hadoop Dataset.

From: A unified framework for detecting point and collective anomalies in operating system logs via collaborative transformers

Anomaly detection technique

Training time (Seconds)

Inference time (Seconds)

Overall

Per-sample

Overall

Per-sample

LSTM36

7.72

0.00112

0.14

0.00002

Attentional BiLSTM63

25.71

0.00372

0.21

0.00003

Convolutional Neural Network46

14.85

0.00215

0.07

0.00001

Transformer80

9.05

0.00131

0.09

0.00001

pylogsentiment38

2615.29

0.01138

94.78

0.00262

CoLog

2479.02

0.01079

124.00

0.00343