README.md

CatBoost tutorials

Basic

It's better to start CatBoost exploring from this basic tutorials.

Python

• 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

• 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

• Command Line Tutorial
  • This tutorial shows how to train and apply model with the command line tool.

Classification

• Classification Tutorial
  • Here is an example for CatBoost to solve binary classification and multi-classification problems.

Ranking

• Ranking Tutorial
  • CatBoost is learning to rank on Microsoft dataset (msrank).

Feature selection

• Feature selection Tutorial
  • This tutorial shows how to make feature evaluation with CatBoost and explore learning rate.

Model analysis

• Object Importance Tutorial

  • 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.

• SHAP Values Tutorial

  • This tutorial shows how to use SHAP python-package to get and visualize feature importances.

• Export CatBoost Model in JSON format Tutorial

  • This tutorial shows how to save catboost model in JSON format and apply it.

• Visualization of CatBoost decision trees tutorial

  • This tutorial shows how to visualize catboost decision trees.

• Feature statistics tutorial

  • This tutorial shows how to calculate feature statistics for catboost model.

Custom loss

• Custom Metrics Tutorial
  • This tutorial shows how to add custom per-object metrics.

Apply model

• CatBoost CoreML Tutorial

  • Explore this tutorial to learn how to convert CatBoost model to CoreML format and use it on any iOS device.

• Export CatBoost Model as C++ code Tutorial

  • Catboost model could be saved as standalone C++ code.

• Export CatBoost Model as Python code Tutorial

  • Catboost model could be saved as standalone Python code.

• Apply CatBoost model from Java

  • 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

• Apply CatBoost model from Rust

  • 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 to use CatBoost fast appliers

  • 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.

Tools

• Gradient Boosting: CPU vs GPU

  • This is a basic tutorial which shows how to run gradient boosting on CPU and GPU on Google Colaboratory.

• Regression on Gradient Boosting: CPU vs GPU

  • This is a basic tutorial which shows how to run regression on gradient boosting on CPU and GPU on Google Colaboratory.

Competition examples

• Kaggle Paribas Competition Tutorial

  • 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.

• ML Boot Camp V Competition Tutorial

  • This is an actual 7th place solution by Mikhail Pershin. Solution is very simple and is based on CatBoost.

• CatBoost & TensorFlow Tutorial

  • This tutorial shows how to use CatBoost together with TensorFlow on Kaggle Quora Question Pairs competition if you have text as input data.

Events

• PyData Moscow tutorial

  • Tutorial from PyData Moscow, October 13, 2018.

• PyData NYC tutorial

  • Tutorial from PyData New York, October 19, 2018.

• PyData LA tutorial

  • Tutorial from PyData Los Angeles, October 21, 2018.

• PyData Moscow tutorial

  • Tutorial from PyData Moscow, April 27, 2019.

• PyData London tutorial

  • Tutorial from PyData London, June 15, 2019.

• PyData Boston tutorial

  • Tutorial from PyData Boston, April 30, 2019.

Tutorials in Russian

• Find tutorials in Russian on the separate page.