ATM Transaction Forecasting on Daily transaction data (Anomaly-Affected Data - AD)
30.69SMAPE (%)FT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| FTBackbone=TimesNet, Forecast Horizon=14-day, Training Strategy=Fine-Tuning (adapted on anomalous samples)2025.12 | 30.69 | -7.22 | |
| WECABackbone=TimesNet, Forecast Horizon=14-day, Training Strategy=Weighted-Contrastive Anomaly-Aware Adaptation2025.12 | 31.78 | -6.13 | |
| CL-ILBackbone=TimesNet, Forecast Horizon=14-day, Training Strategy=Instance Contrastive Learning (contrastive pretraining with w=1)2025.12 | 33.09 | -4.82 | |
| NTBackbone=TimesNet, Forecast Horizon=14-day, Training Strategy=Normally trained (trained on normal data only)2025.12 | 37.91 | — |