CatBoost is well covered with educational materials for both novice and advanced machine learners and data scientists.
This tutorial gives a short introduction to CatBoost and showcases its' functionality in Jupyter Notebook.
Note that the links to the materials at the start of the presentation no longer work, you can find the Jupyter notebook here instead.
Getting started tutorials
These Python tutorials show how to start working with CatBoost.
Perform the following steps to use them:
Download the tutorials using one of the following methods:
- Click the Download button on the github page
- Clone the whole repository using the following command:
git clone https://github.com/catboost/tutorials
Run Jupyter Notebook in the directory with the required
CatBoost on GPU
This tutorial shows how to run CatBoost on GPU with Google Colaboratory.
Tutorials in the CatBoost repository
The CatBoost repository contains several tutorials on various topics, including but no limited to:
- how to apply the model
- how to use custom losses
- how to train a ranking model
- how to perform hyperparameter search
Check out a free part of the Introduction to Competitive Data Science course. The assignment helps to explore all basic functions and implementation features of the CatBoost Python package and understand how to win a Data Science Competition. (in Russian)
Applying CatBoost models in ClickHouse
The ClickHouse documentation contains a tutorial on applying a CatBoost model in ClickHouse.