Classification on PAMAP2
66.791NN AccuracyShiFT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| ShiFTTemperature (tau)=1.0, Evaluation protocol=1NN classifier, Reference set sampling=balanced 1% sample, Number of random seeds=52026.06 | 66.79 | — | |
| InfoTSEvaluation protocol=1NN classifier, Reference set sampling=balanced 1% sample, Number of random seeds=52026.06 | 59.11 | — | |
| ShiFTTemperature (tau)=0.1, Evaluation protocol=1NN classifier, Reference set sampling=balanced 1% sample, Number of random seeds=52026.06 | 58.87 | — | |
| TS2VecEvaluation protocol=1NN classifier, Reference set sampling=balanced 1% sample, Number of random seeds=52026.06 | 57.9 | — | |
| SimCLREvaluation protocol=1NN classifier, Reference set sampling=balanced 1% sample, Number of random seeds=52026.06 | 56.9 | — | |
| Rand Init.Evaluation protocol=1NN classifier, Reference set sampling=balanced 1% sample, Number of random seeds=52026.06 | 55.78 | — | |
| SimMTMEvaluation protocol=1NN classifier, Reference set sampling=balanced 1% sample, Number of random seeds=52026.06 | 54.75 | — | |
| BaselineModality=Time-series, Evaluation Protocol=In-domain linear evaluation, Pre-training=Random initialization2023.10 | — | 69.8 | |
| e-MixModality=Time-series, Evaluation Protocol=In-domain linear evaluation, Framework=Self-supervised learning2023.10 | — | 80.1 | |
| MAEModality=Time-series, Evaluation Protocol=In-domain linear evaluation, Framework=Self-supervised learning2023.10 | — | 85.3 | |
| MetaMAEModality=Time-series, Evaluation Protocol=In-domain linear evaluation, Framework=Self-supervised learning2023.10 | — | 89.3 | |
| ShEDModality=Time-series, Evaluation Protocol=In-domain linear evaluation, Framework=Self-supervised learning2023.10 | — | 85.2 |