Table 8 The fitting errors of the autoencoders in the three CLeaR instances and the baseline model. ae refers to the autoencoder
From: CLeaR: An adaptive continual learning framework for regression tasks
Dataset | Fitting error (e-2) | |||
|---|---|---|---|---|
| Â | Instance\(\phantom {\dot {i}\!}_{\mathbf {A}_{{ae}}}\) | Instance\(\phantom {\dot {i}\!}_{\mathbf {B}_{{ae}}}\) | Instance\(\phantom {\dot {i}\!}_{\mathbf {C}_{{ae}}}\) | Baseline ae |
WF 1 | 2.098 | 0.369 | 0.502 | 1.505 |
WF 2 | 2.068 | 0.348 | 0.440 | 1.579 |
WF 3 | 2.292 | 0.423 | 0.513 | 1.634 |
WF 4 | 2.669 | 0.988 | 0.863 | 1.449 |
WF 5 | 1.421 | 0.516 | 0.471 | 1.549 |
WF 6 | 2.456 | 1.054 | 0.761 | 1.065 |
WF 7 | 2.129 | 0.647 | 0.931 | 1.441 |
WF 8 | 2.149 | 0.434 | 0.662 | 1.448 |
WF 9 | 2.370 | 0.773 | 0.793 | 1.340 |
WF 10 | 1.996 | 0.678 | 0.411 | 1.406 |
Mean | 2.165 | 0.623 | 0.635 | 1.442 |