Table 4 The number of parameters and time consumption (training ratio is 50%)
From: A electricity theft detection method through contrastive learning in smart grid
Models | Parameters (\(10^6\)) | Training time (s) | Testing time (s) |
|---|---|---|---|
WDCNN | 0.896 | 1.74 | 1.03 |
HybridAttn | 51.062 | 14.94 | 4.29 |
HORLN | 3.427 | 3.81 | 1.11 |
Proposed method | 1.065 | 8.58 | 1.54 |