Table 6. The ten top performing models when ranked by validation error.
Training error | NRMSE | Regressor | Test error | Dataset |
---|---|---|---|---|
89.38 | 0.26 | RR | 46.23 | PHYLUM_NONCURATED_JOINT |
88.93 | 0.26 | ENR | 47.82 | PHYLUM_NONCURATED_JOINT |
74.55 | 0.30 | LR | 49.31 | SPECIES_NONCURATED_JOINT |
90.03 | 0.27 | SVR | 49.80 | PHYLUM_NONCURATED_JOINT |
73.62 | 0.28 | SVR | 50.52 | PHYLUM_CURATED_JOINT |
67.57 | 0.32 | RR | 51.14 | SPECIES_NONCURATED_JOINT |
67.40 | 0.32 | ENR | 51.63 | SPECIES_NONCURATED_JOINT |
89.51 | 0.29 | LR | 51.71 | PHYLUM_NONCURATED_JOINT |
78.66 | 0.35 | RR | 51.94 | ORDER_CURATED_JOINT |
74.66 | 0.30 | KNR | 52.17 | CLASS_CURATED_JOINT |