Research map / Label-efficient and robust learning / Recognition and long-tail
Binary Split Categorical Feature with Mean Absolute Error Criteria in CART
Peng Yu, Yike Chen, Chao Xu, Albert Bifet, Jesse Read
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 the exact O(n log n + k log k log n) MAE split instead of LightGBM subsampled heuristic | MAE categorical binary split · Predict Droughts / TS | time (s): 5.40698 -> 0.948483 | Improved |
| use the exact MAE split instead of LightGBM | MAE categorical binary split · wine / fixed acidity | time (s): 0.012946 -> 0.000178 | Improved |
| use the exact MAE split instead of LightGBM | MAE categorical binary split · diamonds / carat | time (s): 0.021328 -> 0.016229 | Improved |
| use the exact MAE split instead of LightGBM | MAE categorical binary split · delays_zurich_transport / stop_id | time (s): 1.29020 -> 0.370960 | Improved |
| use exact MAE split instead of LightGBM subsampled split | MAE categorical binary split · diamonds / carat | relative accuracy: 0.6193 | Improved |
| use exact MAE split instead of LightGBM subsampled split | MAE categorical binary split · delays_zurich_transport / windspeed_avg | relative accuracy: 0.9996 | No clear effect |
| use the median-based numerical encoding instead of an exact split | MAE splitting · constructed counter-example | split objective: n+2ε vs n/2+1 | Hurt |
Closest papers by meaning
- Sparse Learning with CART (NeurIPS 2020 · similarity 0.85)
- A benchmark of categorical encoders for binary classification (NeurIPS 2023 · similarity 0.84)
- Cross-Validated Variable Selection in Tree-Based Methods Improves Predictive Performance (TPAMI 2016 · similarity 0.83)
- Random Forests, Decision Trees, and Categorical Predictors: The "Absent Levels" Problem (JMLR 2018 · similarity 0.81)
- Measures of diversity and space-filling designs for categorical data (ICML 2024 · similarity 0.81)
- Symbolic Regression Enhanced Decision Trees for Classification Tasks (AAAI 2024 · similarity 0.80)
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.
