catboost.ai Open in urlscan Pro
2a02:6b8::1b  Public Scan

Form analysis
0 forms found in the DOM

Text Content

DocumentationGitHubNewsBenchmarksYour FeedbackContacts
 * 
 * 

CatBoost is a high-performance open source library for gradient boosting on
decision trees

How to installTutorials


FEATURES

1
Great quality without parameter tuning
Reduce time spent on parameter tuning, because CatBoost provides great results
with default parameters

2
Categorical features support
Improve your training results with CatBoost that allows you to use non-numeric
factors, instead of having to pre-process your data or spend time and effort
turning it to numbers.

3
Fast and scalable GPU version
Train your model on a fast implementation of gradient-boosting algorithm for
GPU. Use a multi-card configuration for large datasets.

4
Improved accuracy
Reduce overfitting when constructing your models with a novel gradient-boosting
scheme.

5
Fast prediction
Apply your trained model quickly and efficiently even to latency-critical tasks
using CatBoost's model applier



ABOUT

CatBoost is an algorithm for gradient boosting on decision trees. It is
developed by Yandex researchers and engineers, and is used for search,
recommendation systems, personal assistant, self-driving cars, weather
prediction and many other tasks at Yandex and in other companies, including
CERN, Cloudflare, Careem taxi. It is in open-source and can be used by anyone.




LATEST NEWS

Cloudflare's Protection Against Bots Employs CatBoostMarch 7, 2019

Training with CatBoost on a CPU/GPU cluster with a dataset of trillion requests
helps to identify malicious bot traffic.

Careem's Destination Prediction Service uses CatBoostFebruary 19, 2019

Careem, the leading ride-hailing platform for the greater Middle East, explained
how CatBoost helps predicting their customer next move.

CatBoost Enables Fast Gradient Boosting on Decision Trees Using GPUsDecember 18,
2018

Gradient boosting benefits from training on huge datasets. In addition, the
technique is efficiently accelerated using GPUs. Read details in this post.

Full news list


BENCHMARKS

 * Quality
 * Learning speed

TunedDefault
CatBoost
LightGBM
XGBoost
H2O
Tuned
Default
Tuned
Default
Tuned
Default
Tuned
Default
Adult
0.26974
0.00%
0.27298
1.21%
0.27602
2.33%
0.28716
6.46%
0.27542
2.11%
0.28009
3.84%
0.27510
1.99%
0.27607
2.35%
Amazon
0.13772
0.00%
0.13811
0.29%
0.16360
18.80%
0.16716
21.38%
0.16327
18.56%
0.16536
20.07%
0.16264
18.10%
0.16950
23.08%
Click prediction
0.39090
0.00%
0.39112
0.06%
0.39633
1.39%
0.39749
1.69%
0.39624
1.37%
0.39764
1.73%
0.39759
1.72%
0.39785
1.78%
KDD appetency
0.07151
0.00%
0.07138
0.19%
0.07179
0.40%
0.07482
4.63%
0.07176
0.35%
0.07466
4.41%
0.07246
1.33%
0.07355
2.86%
KDD churn
0.23129
0.00%
0.23193
0.28%
0.23205
0.33%
0.23565
1.89%
0.23312
0.80%
0.23369
1.04%
0.23275
0.64%
0.23287
0.69%
KDD internet
0.20875
0.00%
0.22021
5.49%
0.22315
6.90%
0.23627
13.19%
0.22532
7.94%
0.23468
12.43%
0.22209
6.40%
0.24023
15.09%
KDD upselling
0.16613
0.00%
0.16674
0.37%
0.16682
0.42%
0.17107
2.98%
0.16632
0.12%
0.16873
1.57%
0.16824
1.28%
0.16981
2.22%
KDD 98
0.19467
0.00%
0.19479
0.07%
0.19576
0.56%
0.19837
1.91%
0.19568
0.52%
0.19795
1.69%
0.19539
0.37%
0.19607
0.72%
Kick prediction
0.28479
0.00%
0.28491
0.05%
0.29566
3.82%
0.29877
4.91%
0.29465
3.47%
0.29816
4.70%
0.29481
3.52%
0.29635
4.06%
Figures in this table represent Logloss values (lower is better) for
Classification mode.
Percentage is metric difference measured against tuned CatBoost results.



CONTACTS

 * Report an issue with CatBoost on GitHub.
 * Ask a question on Stack Overflow with the catboost tag, we monitor this for
   new questions.
 * Join Telegram chat to discuss with real users in English or in Russian.

© 2024 Yandex
 * Twitter