Developed by Yandex researchers and engineers, CatBoost (which stands for categorical boosting) is a gradient boosting algorithm, based on decision trees, which is optimized in handling categorical features without much preprocessing (non-numeric features expressing a quality, such as a color, a brand, or a type). Since in most databases the majority of features are categorical, CatBoost can really boost your results on prediction:

• Website: https://catboost.yandex
• Version at the time of print: 0.8.1.1
• Suggested install command: pip install catboost
• Download page: https://github.com/catboost/catboost