Back to ULR overview · Back to main index
| Metric | Value |
|---|---|
| Test accuracy | 51.94% |
| Test F1 score | 0.5719 |
| Hierarchical loss | 0.95630625 |
| P-adic loss (total) | 161.67792886 |
| P-adic loss (mean) | 0.23196260 |
| Prime base | 71 |
| Number of tags (input features) | 1,640 |
| Non-zero parameters | 1,616 / 354,456 (99.5% sparse) |
| L1 regularization (C) | 1.0000 |
| Training samples | 2,726 |
| Test samples | 697 |
| Agreement | Count | Share | Cost per mistake | Total contribution |
|---|---|---|---|---|
| Exact match | 362 | 51.94% | 0.000000 | 0.000000 |
| p^6 | 2 | 0.29% | 0.000000 | 0.000000 |
| p^5 | 0 | 0.00% | 0.000000 | 0.000000 |
| p^4 | 14 | 2.01% | 0.000000 | 0.000001 |
| p^3 | 31 | 4.45% | 0.000003 | 0.000087 |
| p^2 | 80 | 11.48% | 0.000198 | 0.015870 |
| p^1 | 47 | 6.74% | 0.014085 | 0.661972 |
| p^0 | 161 | 23.10% | 1.000000 | 161.000000 |
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.