It's better to start CatBoost exploring from this basic tutorials.
- Python Tutorial
- This tutorial shows some base cases of using CatBoost, such as model training, cross-validation and predicting, as well as some useful features like early stopping, snapshot support, feature importances and parameters tuning.
- Python Tutorial with task
- There are 17 questions in this tutorial. Try answering all of them, this will help you to learn how to use the library.
- R Tutorial
- This tutorial shows how to convert your data to CatBoost Pool, how to train a model and how to make cross validation and parameter tunning.
- Command Line Tutorial
- This tutorial shows how to train and apply model with the command line tool.
- Classification Tutorial
- Here is an example for CatBoost to solve binary classification and multi-classification problems.
- Ranking Tutorial
- CatBoost is learning to rank on Microsoft dataset (msrank).
- Feature selection Tutorial
- This tutorial shows how to make feature evaluation with CatBoost and explore learning rate.
- This tutorial shows how to evaluate importances of the train objects for test objects, and how to detect broken train objects by using the importance scores.
- This tutorial shows how to use SHAP python-package to get and visualize feature importances.
- This tutorial shows how to save catboost model in JSON format and apply it.
- This tutorial shows how to visualize catboost decision trees.
- This tutorial shows how to calculate feature statistics for catboost model.
- This tutorials shows how to use PredictionDiff feature importances.
- Custom Metrics Tutorial
- This tutorial shows how to add custom per-object metrics.
- Explore this tutorial to learn how to convert CatBoost model to CoreML format and use it on any iOS device.
- Catboost model could be saved as standalone C++ code.
- Catboost model could be saved as standalone Python code.
- Explore how to apply CatBoost model from Java application. If you just want to look at code snippets you can go directly to CatBoost4jPredictionTutorial.java
- Explore how to apply CatBoost model from Rust application. If you just want to look at code snippets you can go directly to main.rs
- Convert LightGBM to CatBoost, save resulting CatBoost model and use CatBoost C++, Python, C# or other applier, which in case of not symmetric trees will be around 7-10 faster than native LightGBM one.
- Note that CatBoost applier with CatBoost models is even faster, because it uses specific fast symmetric trees.
- This is a basic tutorial which shows how to run gradient boosting on CPU and GPU on Google Colaboratory.
- This is a basic tutorial which shows how to run regression on gradient boosting on CPU and GPU on Google Colaboratory.
- This tutorial shows how to get to a 9th place on Kaggle Paribas competition with only few lines of code and training a CatBoost model.
- This is an actual 7th place solution by Mikhail Pershin. Solution is very simple and is based on CatBoost.
- This tutorial shows how to use CatBoost together with TensorFlow on Kaggle Quora Question Pairs competition if you have text as input data.
- Tutorial from PyData Moscow, October 13, 2018.
- Tutorial from PyData New York, October 19, 2018.
- Tutorial from PyData Los Angeles, October 21, 2018.
- Tutorial from PyData Moscow, April 27, 2019.
- Tutorial from PyData London, June 15, 2019.
- Tutorial from PyData Boston, April 30, 2019.
Tutorials in Russian
- Find tutorials in Russian on the separate page.