CatBoost: gradient boosting with categorical features support
About
In this paper we present CatBoost, a new open-sourced gradient boosting library that successfully handles categorical features and outperforms existing publicly available implementations of gradient boosting in terms of quality on a set of popular publicly available datasets. The library has a GPU implementation of learning algorithm and a CPU implementation of scoring algorithm, which are significantly faster than other gradient boosting libraries on ensembles of similar sizes.
Anna Veronika Dorogush, Vasily Ershov, Andrey Gulin• 2018
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Classification | Petfinder (test) | Accuracy38.69 | 16 | |
| Classification | blood-transfusion | AUROC70.9 | 16 | |
| Classification | Adult | ROC-AUC0.93 | 13 | |
| Regression | Crossed Barrel Dataset (80% uniform sampling) | R^20.76 | 10 | |
| Regression | Cogni-e-Spin Dataset (80% uniform sampling) | R^20.55 | 10 | |
| Surrogate Modeling | Crossed Barrel Dataset biased sampling | R^20.55 | 10 | |
| Surrogate Modeling | Cogni-e-Spin Dataset biased sampling | R^20.36 | 10 | |
| Regression | Lattice Dataset (80% uniform sampling) | R^20.95 | 10 | |
| Surrogate Modeling | Lattice Dataset biased sampling | R^20.67 | 10 | |
| Classification | bank-marketing 1461 | AUROC93.7 | 8 |
Showing 10 of 34 rows