Multiclass Classification on Balance UCI
84.4AccuracyAlgorithm 1 (Agn)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Algorithm 1 (Agn)relabeling=fractional, number of weak learners=100, weak learner=VeryFastDecisionTree (depth 1), k=32022.05 | 84.4 | 11.5 | |
| OnlineMBBM (Opt)number of weak learners=100, weak learner=VeryFastDecisionTree (depth 1), k=32022.05 | 79.5 | 57.6 | |
| OCOR (OCO-based online realizable)number of weak learners=100, weak learner=VeryFastDecisionTree (depth 1), k=32022.05 | 78.8 | 4.5 | |
| Adaboost.OLM (Ada)number of weak learners=100, weak learner=VeryFastDecisionTree (depth 1), k=32022.05 | 75.2 | 7.1 |