Research map / Learning theory and algorithms / Prediction and decision losses
Alternating optimization of decision trees, with application to learning sparse oblique trees
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| run TAO alternating optimization on the CART oblique tree | CART oblique tree depth 12 · MNIST test | test error (%): 11.03% -> 5.66% | Helped |
| run TAO on the CART oblique tree | CART oblique tree depth 12 · MNIST train | train error (%): 1.95% -> 0.09% | Helped |
| add strong l1 sparsity penalty (C=0.01) in TAO | TAO oblique tree · MNIST test | test error (%): 5.66% -> 10.19% | Hurt |
| moderately increase l1 sparsity penalty (decrease C) | TAO oblique tree · MNIST test | test error (%) | No clear effect |
| TAO sparse oblique tree vs forest-based nearest neighbor classifiers | TAO · MNIST test | test error (%): 5.69 vs 2.24 (BF) | Hurt |
| TAO single tree vs forest-based classifiers | TAO · protein and seismic test | test error (%): 31.70 vs 44.20 (BF); 27.81 vs 40.60 (BF) | Helped |
| TAO tree instead of forest classifiers | TAO · MNIST test set | inference time (s): 0.18 vs 23.90 (BF) | Helped |
| change TAO node visiting order (top-bottom or alternating instead of bottom-top) (ablation) | TAO · axis-aligned and oblique trees | test error | No clear effect |
| initialize TAO from random tree instead of CART tree (ablation) | TAO · general | test error | No clear effect |
Closest papers by meaning
- Near-Optimal Decision Trees in a SPLIT Second (ICML 2025 · similarity 0.90)
- GradTree: Learning Axis-Aligned Decision Trees with Gradient Descent (AAAI 2024 · similarity 0.89)
- Multistage Defer Trees for Hybrid Interpretability: If at First You Can't Succeed, Tree Again (NeurIPS 2026 · similarity 0.88)
- Optimal Decision Tree Pruning Revisited: Algorithms and Complexity (ICML 2025 · similarity 0.88)
- Branches: Efficiently Seeking Optimal Sparse Decision Trees via AO* (ICML 2025 · similarity 0.88)
- Learning Small Decision Trees with Few Outliers: A Parameterized Perspective (AAAI 2024 · similarity 0.88)
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.