Back to ULR overview · Back to main index
| Metric | Value |
|---|---|
| Test accuracy | 59.16% |
| Test F1 score | 0.6240 |
| Hierarchical loss | 0.92283269 |
| P-adic loss (total) | 231.47657724 |
| P-adic loss (mean) | 0.25691074 |
| Prime base | 79 |
| Number of tags (input features) | 6,936 |
| Non-zero parameters | 2,706 / 2,421,013 (99.9% sparse) |
| L1 regularization (C) | 1.0000 |
| Training samples | 3,679 |
| Test samples | 901 |
| Agreement | Count | Share | Cost per mistake | Total contribution |
|---|---|---|---|---|
| Exact match | 537 | 59.60% | 0.000000 | 0.000000 |
| p^5 | 7 | 0.78% | 0.000000 | 0.000000 |
| p^4 | 13 | 1.44% | 0.000000 | 0.000000 |
| p^3 | 25 | 2.77% | 0.000002 | 0.000051 |
| p^2 | 51 | 5.66% | 0.000160 | 0.008172 |
| p^1 | 37 | 4.11% | 0.012658 | 0.468354 |
| p^0 | 231 | 25.64% | 1.000000 | 231.000000 |
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 (2,706) indicates how many coefficients the model actually uses.