CatBoost: unbiased boosting with categorical features

By Liudmila Ostroumova, Gleb Gusev, Aleksandr Vorobev, Anna Veronika Dorogush, Andrey Gulin. Neural Information Processing Systems, 2017

6,070 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/ee0a0f04d45f86bf50b24d7258e884725fcaa621

How researchers see this paper

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

Citations in the corpus

CatBoost for big data: an interdisciplinary review (2020, methodology): “In the seminal paper on CatBoost, “Catboost: unbiased boosting with categorical features” [2], Prokhorenkova et al.”
Machine Unlearning for Random Forests (2020, background): “One could even investigate the possibility of extending DaRE forests to boosted trees (Chen & Guestrin, 2016; Ke et al., 2017; Prokhorenkova et al., 2018).”
Forecasting gold price with the XGBoost algorithm and SHAP interaction values (2021, unknown): “According to Prokhorenkova et al. (2018), CatBoost uses ordered boosting and an innovative algorithm for processing categorical features.”
Strength prediction of concrete-filled steel tubular columns using Categorical Gradient Boosting algorithm (2021, background): “CatBoost proposed by Yandex [42] is a new type of GBM that can handle a categorical features well.”
On Embeddings for Numerical Features in Tabular Deep Learning (2022, unknown): “However, when systematically evaluated, these models do not consistently outperform the ensembles of decision trees, such as GBDT (Gradient Boosting Decision Tree) [7, 19, 32], which are typically the top-choice in various ML competitions [13, 36].”
Boosting methods for multi-class imbalanced data classification: an experimental review (2020, background): “CatBoost has both CPU and GPU implementations which accelerates the training process [50].”
A comparative analysis of gradient boosting algorithms (2019, unknown): “The performance of CatBoost and LightGBM is promising specially in generalization accuracy for the former (Prokhorenkova et al. 2018) and in training speed for the latter (Ke et al.”
Transformers Can Do Bayesian Inference (2021, methodology): “Catboost additionally received information on which features are categorical.”
ADBench: Anomaly Detection Benchmark (2022, methodology): “…(CV) and Transformer [171] in the natural language processing (NLP) domain—we adapt ResNet and FTTransformer models [56] for tabular AD in the proposed ADBench; and ( iv ) ensemble learning methods like LightGBM [74], XGBoost [29], and CatBoost [138] that have shown effectiveness in AD tasks [170].”
DART: Data Addition and Removal Trees (2020, unknown): “One could even investigate the possibility of extending DART to boosted trees (Chen and Guestrin 2016; Ke, Meng et al. 2017; Prokhorenkova, Gusev et al. 2018).”
TabDDPM: Modelling Tabular Data with Diffusion Models (2022, methodology): “Tables 4, 5 show that the absolute values of classification/regression performance are much lower for the first protocol, i.e., weak classifiers/regressors are substantially inferior to CatBoost on the considered benchmarks.”
When Do Neural Nets Outperform Boosted Trees on Tabular Data? (2023, methodology): “CatBoost (Categorical Boosting) [48] introduces ordered boosting, a new method for handling categorical features, as well as better methods for handling missing values and outliers.”
Large Language Models Are Zero-Shot Time Series Forecasters (2023, background): “• Monash [18]: The Monash forecasting archive contains 30 publicly available datasets along with baseline numbers for 12 forecasting models, including simple exponential smooth (e.g. ETS [26]), gradient boosting (e.g. CatBoost [39]) and deep learning models (e.g. DeepAR [40], WaveNet [35]).”
Deep Neural Networks and Tabular Data: A Survey (2021, methodology): “This approach is also used in the CatBoost framework [71], a state-of-the-art machine learning library for heterogeneous tabular data based on the gradient boosting algorithm [95].”