ULR Fold 2

Back to ULR overview · Back to main index

Fold metrics

MetricValue
Test accuracy46.56%
Test F1 score0.5092
Hierarchical loss0.89613142
P-adic loss (total)738.75373481
P-adic loss (mean)0.27340997
Prime base79
Number of tags (input features)4,981
Non-zero parameters6,620 / 3,318,012 (99.8% sparse)
L1 regularization (C)1.0000
Training samples10,857
Test samples2,702

P-adic loss breakdown

AgreementCountShareCost per mistakeTotal contribution
Exact match1,25846.56%0.0000000.000000
p^620.07%0.0000000.000000
p^500.00%0.0000000.000000
p^4521.92%0.0000000.000001
p^31626.00%0.0000020.000329
p^227810.29%0.0001600.044544
p^12147.92%0.0126582.708861
p^073627.24%1.000000736.000000

About L1 regularization

L1 (Lasso) regularization promotes sparsity by driving many coefficients to exactly zero. This model uses ALL available tags (4,981) but L1 regularization selects which features are actually used. The number of non-zero parameters (6,620) indicates how many coefficients the model actually uses.