Back to ULR overview · Back to main index
| Metric | Value |
|---|---|
| Test accuracy | 58.37% |
| Test F1 score | 0.6140 |
| Hierarchical loss | 0.91859346 |
| P-adic loss (total) | 244.52448014 |
| P-adic loss (mean) | 0.27661140 |
| Prime base | 79 |
| Number of tags (input features) | 6,936 |
| Non-zero parameters | 2,654 / 2,421,013 (99.9% sparse) |
| L1 regularization (C) | 1.0000 |
| Training samples | 3,696 |
| Test samples | 884 |
| Agreement | Count | Share | Cost per mistake | Total contribution |
|---|---|---|---|---|
| Exact match | 525 | 59.39% | 0.000000 | 0.000000 |
| p^5 | 5 | 0.57% | 0.000000 | 0.000000 |
| p^4 | 13 | 1.47% | 0.000000 | 0.000000 |
| p^3 | 22 | 2.49% | 0.000002 | 0.000045 |
| p^2 | 34 | 3.85% | 0.000160 | 0.005448 |
| p^1 | 41 | 4.64% | 0.012658 | 0.518987 |
| p^0 | 244 | 27.60% | 1.000000 | 244.000000 |
L1 (Lasso) regularization promotes sparsity by driving many coefficients to exactly zero. This model uses ALL available tags (6,936) but L1 regularization selects which features are actually used. The number of non-zero parameters (2,654) indicates how many coefficients the model actually uses.