Times are summed over the 8 CV folds (fit + predict/score). They come from per-fold timings in the scores CSV when available, otherwise from the run log. Hardware and parallel load differ across runs (CPU vs GPU), so times are indicative rather than a controlled benchmark. The horizontal axis is log-scaled. MAPE uses the median prediction. Pinball loss is averaged over the 5%, 50% and 95% quantiles and all horizons; models without quantiles are omitted from the pinball figure. Lower is better on both axes. Table values are means across the 8 folds.
| model | fit + predict (s) | MAPE (%) | pinball loss (MW) |
|---|---|---|---|
| tabicl_n8_full | 523 | 1.654 | 217.66 |
| kumo_forecast | 558 | 3.776 | - |
| hgb_mean | 692 | 1.672 | - |
| kumo_tabular | 1,240 | 1.281 | 174.57 |
| hgb_quantile_q7 | 2,387 | 1.675 | 230.53 |
| chronos2_ctx4000 | 3,090 | 1.635 | 208.88 |
| hgb_quantile_recal_q7 | 11,002 | 1.742 | 230.32 |
MAPE is relative to hgb_mean;
pinball loss is relative to hgb_quantile_q7.
The baseline is shown at zero difference. Large markers: pooled mean difference over all test rows, with a 95% moving-block bootstrap interval (14-day blocks over daily differences; conditional on the fitted models, it ignores training randomness). Small dots: mean difference in each of the 8 CV folds (one dot per fold; hover shows the fold). Folds are not independent, so no t-interval across folds is shown. Dashed colored lines
mark the Pareto frontiers.
Means across the 8 CV folds, shown without error bars. Dashed colored lines mark the Pareto frontiers.