BORN AGAIN TREES

By Leo Breiman, Nong Shang. 1996

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Yoshua Bengio
2. Dacheng Tao
3. Luc Van Gool
4. Trevor Darrell
5. Bernt Schiele
6. Yi Yang
7. Sergey Levine
8. Chunhua Shen

Citations in the corpus

• Interpreting tree ensembles with inTrees (2014, methodology): “To gain interpretability for tree ensembles, methods were proposed to approximate tree ensembles by single decision trees [7,12,20].”
• Dreaming to Distill: Data-Free Knowledge Transfer via DeepInversion (2019, background): “Breiman and Shang first introduced this concept in learning a single decision tree to approximate the outputs of multiple decision trees [4].”
• Relational Knowledge Distillation (2019, background): “Breiman and Shang [3] first proposed to learn single-tree models that approximate the performance of multiple-tree models and provide better interpretability.”
• Interpreting Blackbox Models via Model Extraction (2017, methodology): “There has been work using model compression [5] to learn decision trees [4, 15], but they use rejection sampling, whereas our active sampling strategy directly targets paths most in need of additional data, thereby substantially improving accuracy.”
• Intra-class Feature Variation Distillation for Semantic Segmentation (2020, result): “Similar ideas can also be found in [5,7,3].”
• Ensemble learning: A survey (2018, methodology): “The final decision tree is trained using the manufactured data labels (Breiman & Shang, 1996).”
• Born Again Neural Networks (2018, background): “Previously, in Born-Again Trees Breiman & Shang (1996) pioneered this idea, learning single trees that match the performance of multiple-tree predictors.”
• Born-Again Tree Ensembles (2020, background): “In their seminal work about born-again trees, Breiman & Shang (1996) were the first to introduce a thinning problem that aimed at replacing a tree ensemble by a newly constructed simpler classifier.”
• Fast, Accurate, and Simple Models for Tabular Data via Augmented Distillation (2020, methodology): “For tabular data, Breiman and Shang [15] considered distilling models into single decision trees, but this often unacceptably harms accuracy.”
• Making Tree Ensembles Interpretable: A Bayesian Model Selection Approach (2016, methodology): “There are a few seminal studies on interpreting tree ensembles: Born Again Trees (BATrees) [17]; interpretable Trees (inTrees) [18]; and Node Harvest [19].”
• Locally Interpretable Models and Effects based on Supervised Partitioning (LIME-SUP) (2018, background): “One of the earliest implementations that we know of is the concept of “born again trees” (Breiman & Shang, 1997).”