Research map / Learning theory and algorithms / Prediction and decision losses
Making Decision Trees Feasible in Ultrahigh Feature and Label Dimensions
Weiwei Liu, Ivor W. Tsang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add cross-validated pruning (ablation) | SBT · epsilon | test error rate (%): 17.23 -> 12.45 | Improved |
| add cross-validated pruning (ablation) | SBT · rcv | test error rate (%): 6.19 -> 3.85 | Improved |
| use SBT instead of CART-LC | tree classifier · epsilon | test error rate (%): 49.40 -> 12.45 | Improved |
| use SBT (feature-selected sparse hyperplanes) instead of l2-SVM | tree classifier · colon | test error rate (%): 20.00 -> 13.33 | Improved |
| use SBT instead of LDKL | tree classifier · epsilon | test error rate (%): 15.36 -> 12.45 | Improved |
| use SBT instead of GBFS for feature selection training | feature selection · rcv, 100 selected features | training time (s): 874.2s -> 5.9s | Improved |
| use SBT instead of LDKL training | tree classifier · rcv | training time (s): 8019.3 -> 41.9 | Improved |
| add tree structure to budgeted feature selection (SBT instead of FGM) | feature selection · pcmac, gisette, epsilon, rcv | test error vs selected features | Improved |
| add label-frequency-ordered annotation (AT instead of PT) | multi-label tree · EUR-Lex(ed) | F1: 0.274 -> 0.341 | Improved |
| use AT instead of PT | multi-label tree · amazon | F1: 0.0166 -> 0.0348 | Improved |
| use AT instead of PD-Sparse | multi-label · amazon | F1: 0.0151 -> 0.0348 | Improved |
| use AT instead of FastXML | multi-label tree · wiki | F1: 0.1272 -> 0.1657 | Improved |
Closest papers by meaning
- Symbolic Regression Enhanced Decision Trees for Classification Tasks (AAAI 2024 · similarity 0.88)
- Multistage Defer Trees for Hybrid Interpretability: If at First You Can't Succeed, Tree Again (NeurIPS 2026 · similarity 0.87)
- Branches: Efficiently Seeking Optimal Sparse Decision Trees via AO* (ICML 2025 · similarity 0.87)
- Learning Small Decision Trees with Few Outliers: A Parameterized Perspective (AAAI 2024 · similarity 0.87)
- Near-Optimal Decision Trees in a SPLIT Second (ICML 2025 · similarity 0.86)
- GradTree: Learning Axis-Aligned Decision Trees with Gradient Descent (AAAI 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
