Research map / Learning theory and algorithms / Prediction and decision losses
Maximum Margin Interval Trees
Alexandre Drouin, Toby Hocking, Francois Laviolette
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 |
|---|---|---|---|
| use nonlinear MMIT trees instead of L1-Linear | Interval regression · Simulated sin and abs | test MSE | Improved |
| use L1-Linear instead of MMIT | Interval regression · Simulated linear | test MSE | Improved |
| use squared hinge loss instead of linear hinge loss (ablation) | MMIT · Histone | test MSE | Improved |
| use squared hinge loss instead of linear hinge loss (ablation) | MMIT · Across reported datasets | test MSE | Mixed |
| use an interval-aware margin loss instead of squared loss on interval limits | Interval-CART · Histone | test MSE | Improved |
| use margin-based tree models instead of other models | Interval regression · Servo | test MSE | Improved |
| use margin-based models instead of Constant or TransfoTree | Interval regression · Histone | test MSE | Improved |
| use MMIT instead of L1-Linear | Interval regression · Neuroblastoma | test MSE | No clear effect |
| use MMIT instead of Constant | Interval regression · Triazines | test MSE | No clear effect |
| use MMIT instead of TransfoTree, L1-Linear, or Constant | Interval regression · Simulated linear | test MSE | Mixed |
Closest papers by meaning
- Piecewise Constant and Linear Regression Trees: An Optimal Dynamic Programming Approach (ICML 2024 · similarity 0.86)
- Regression Trees Know Calculus (NeurIPS 2025 · similarity 0.84)
- Optimal Classification Trees for Continuous Feature Data Using Dynamic Programming with Branch-and-Bound (AAAI 2025 · similarity 0.84)
- Boosting Methods for Interval-censored Data with Regression and Classification (ICLR 2025 · similarity 0.83)
- Learning from Interval Targets (NeurIPS 2025 · similarity 0.83)
- Minimax Rates and Spectral Distillation for Tree Ensembles (NeurIPS 2026 · similarity 0.83)
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.
