ULR Fold 2

Back to ULR overview · Back to main index

Fold metrics

MetricValue
Test accuracy51.94%
Test F1 score0.5719
Hierarchical loss0.95630625
P-adic loss (total)161.67792886
P-adic loss (mean)0.23196260
Prime base71
Number of tags (input features)1,640
Non-zero parameters1,616 / 354,456 (99.5% sparse)
L1 regularization (C)1.0000
Training samples2,726
Test samples697

P-adic loss breakdown

AgreementCountShareCost per mistakeTotal contribution
Exact match36251.94%0.0000000.000000
p^620.29%0.0000000.000000
p^500.00%0.0000000.000000
p^4142.01%0.0000000.000001
p^3314.45%0.0000030.000087
p^28011.48%0.0001980.015870
p^1476.74%0.0140850.661972
p^016123.10%1.000000161.000000

About L1 regularization

L1 (Lasso) regularization promotes sparsity by driving many coefficients to exactly zero. This model uses ALL available tags (1,640) but L1 regularization selects which features are actually used. The number of non-zero parameters (1,616) indicates how many coefficients the model actually uses.