Time Series Failure Prediction on Real-world ATM Machine Failure Prediction (modified splits)
98.5AccuracyContextFlow++
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| ContextFlow++context_encoder=Embed. determ.2024.06 | 98.5 | 95.9 | 99.6 | 97.1 | 92.6 | 92.6 | |
| ContextFlow++context_encoder=One-hot variational2024.06 | 98.3 | 95 | 99.4 | 97 | 91.9 | 90.9 | |
| ContextFlow++context_encoder=Integer arg max2024.06 | 98.1 | 95.1 | 99.4 | 96 | 90.9 | 91.2 | |
| ContextFlows Generalist Flowmodel_type=reference flow-based generalist2024.06 | 97.2 | 91.5 | 98.7 | 92.7 | 86.3 | 84.4 | |
| XGBoost2024.06 | 96.7 | 85.3 | 85.3 | 71 | 81.5 | 70.9 | |
| Embed. determ. Lu and Huangcontext_conditioning=Embed. determ.2024.06 | 96.7 | 91.9 | 99 | 93.2 | 84.3 | 85.9 | |
| random forests2024.06 | 94.39 | 73.71 | 73.71 | 51.73 | 63.68 | 47.59 | |
| HYDRA2024.06 | 81.3 | 74.1 | 74.1 | 23.8 | 42 | 65 |