ULR Fold 0

Back to ULR overview · Back to main index

Fold metrics

MetricValue
Test accuracy54.49%
Test F1 score0.5829
Hierarchical loss0.95862737
P-adic loss (total)141.57854366
P-adic loss (mean)0.21549246
Prime base71
Number of tags (input features)1,640
Non-zero parameters1,650 / 357,738 (99.5% sparse)
L1 regularization (C)1.0000
Training samples2,766
Test samples657

P-adic loss breakdown

AgreementCountShareCost per mistakeTotal contribution
Exact match35854.49%0.0000000.000000
p^610.15%0.0000000.000000
p^500.00%0.0000000.000000
p^4101.52%0.0000000.000000
p^3314.72%0.0000030.000087
p^27611.57%0.0001980.015076
p^1406.09%0.0140850.563380
p^014121.46%1.000000141.000000

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 (1,650) indicates how many coefficients the model actually uses.