Mean Multicalibration
3Rate in ECEProposed lower bound
Evaluation Results
| Method | Links | |
|---|---|---|
| Proposed lower boundGuarantee=Batch lower bound for multicalibration in mean regression and for more general properties, Randomized=true, Bound Type=Lower Bounds2026.04 | 3 | |
| NRRX25Guarantee=Online bucketed mean multicalibration, followed by rounding and Proposition 39, Randomized=true, Bound Type=Upper Bounds2026.04 | 3 | |
| GJN+22Guarantee=Online (α, n)-mean multicalibration on n buckets, plus online-to-batch, Randomized=true, Bound Type=Upper Bounds2026.04 | 4 | |
| HJZ23Guarantee=Batch (G, ε, λ) bucketed multicalibration in L∞, Randomized=true, Bound Type=Upper Bounds2026.04 | 4 | |
| GT25Guarantee=Batch lower bound for calibrated multiaccuracy / ECE in mean regression (for deterministic predictors), Randomized=false, Bound Type=Lower Bounds2026.04 | 5 | |
| HJKRR18Guarantee=Batch (G, α)-multicalibration with discarded mass and a minimum-group-mass assumption, Randomized=false, Bound Type=Upper Bounds2026.04 | 7 | |
| GKSZ22Guarantee=Full multicalibration with interval basis, specialized back to mean ECE, Randomized=false, Bound Type=Upper Bounds2026.04 | 8 | |
| GHHK+23Guarantee=Weighted L2-multicalibration from L2-boosting, Randomized=false, Bound Type=Upper Bounds2026.04 | 10 |