Time Series Forecasting on Job Posting Skill Demand seq_len=4 normalized skill proportions
0MAEFEDformer
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| FEDformerModel family=Transformer / Hybrid, seq_len=4, pred_len=32026.05 | 0 | 0.0001 | 56.29 | 11.03 | |
| ReformerModel family=Transformer / Hybrid, seq_len=4, pred_len=32026.05 | 0 | 0.0001 | 55.8 | 12.36 | |
| TransformerModel family=Transformer / Hybrid, seq_len=4, pred_len=32026.05 | 0 | 0.0001 | 56.07 | 14.05 | |
| InformerModel family=Transformer / Hybrid, seq_len=4, pred_len=32026.05 | 0 | 0.0001 | 56.88 | 11.8 | |
| AutoformerModel family=Transformer / Hybrid, seq_len=4, pred_len=32026.05 | 0 | 0.0001 | 58.35 | 15.64 | |
| Nonstationary_TransformerModel family=Transformer / Hybrid, seq_len=4, pred_len=32026.05 | 0 | 0.0001 | 65.2 | 25.4 | |
| CrossformerModel family=Transformer / Hybrid, seq_len=4, pred_len=32026.05 | 0.0001 | 0.0002 | 63.45 | 30.58 | |
| LSTMModel family=Recurrent / Convolutional, seq_len=4, pred_len=32026.05 | 0.0001 | 0.0002 | 64.7 | 30.86 | |
| TSMixerModel family=Decomposition / Linear, seq_len=4, pred_len=32026.05 | 0.0001 | 0.0002 | 67.86 | 30.55 | |
| CHGHModel family=Decomposition / Linear, seq_len=4, pred_len=32026.05 | 0.0001 | 0.0002 | 66.58 | 32.57 | |
| FreTSModel family=Transformer / Hybrid, seq_len=4, pred_len=32026.05 | 0.0001 | 0.0002 | 65.15 | 32.28 | |
| KoopaModel family=Transformer / Hybrid, seq_len=4, pred_len=32026.05 | 0.0001 | 0.0002 | 72.79 | 40.82 | |
| TiDEModel family=Decomposition / Linear, seq_len=4, pred_len=32026.05 | 0.0001 | 0.0002 | 71.14 | 37.96 | |
| FiLMModel family=Decomposition / Linear, seq_len=4, pred_len=32026.05 | 0.0001 | 0.0002 | 68.95 | 37.53 | |
| DLinearModel family=Decomposition / Linear, seq_len=4, pred_len=32026.05 | 0.0001 | 0.0003 | 82.74 | 49.63 |