LightGBM is a gradient boosting framework that uses tree based learning algorithms. It is designed to be distributed and efficient with the following advantages:
- Faster training speed and higher efficiency.
- Lower memory usage.
- Better accuracy.
- Support of parallel and GPU learning.
- Capable of handling large-scale data.
For further details, please refer to Features.
Benefitting from these advantages, LightGBM is being widely-used in many winning solutions of machine learning competitions.
Comparison experiments on public datasets show that LightGBM can outperform existing boosting frameworks on both efficiency and accuracy, with significantly lower memory consumption. What's more, parallel experiments show that LightGBM can achieve a linear speed-up by using multiple machines for training in specific settings.
Our primary documentation is at https://lightgbm.readthedocs.io/ and is generated from this repository. If you are new to LightGBM, follow the installation instructions on that site.
Next you may want to read:
- Examples showing command line usage of common tasks.
- Features and algorithms supported by LightGBM.
- Parameters is an exhaustive list of customization you can make.
- Parallel Learning and GPU Learning can speed up computation.
- Laurae++ interactive documentation is a detailed guide for hyperparameters.
- Optuna Hyperparameter Tuner provides automated tuning for LightGBM hyperparameters.
Documentation for contributors:
- How we update readthedocs.io.
- Check out the Development Guide.
Please refer to changelogs at GitHub releases page.
Some old update logs are available at Key Events page.
Optuna (hyperparameter optimization framework): https://github.com/optuna/optuna
Julia-package: https://github.com/IQVIA-ML/LightGBM.jl
JPMML (Java PMML converter): https://github.com/jpmml/jpmml-lightgbm
Treelite (model compiler for efficient deployment): https://github.com/dmlc/treelite
cuML Forest Inference Library (GPU-accelerated inference): https://github.com/rapidsai/cuml
m2cgen (model appliers for various languages): https://github.com/BayesWitnesses/m2cgen
leaves (Go model applier): https://github.com/dmitryikh/leaves
ONNXMLTools (ONNX converter): https://github.com/onnx/onnxmltools
SHAP (model output explainer): https://github.com/slundberg/shap
MMLSpark (LightGBM on Spark): https://github.com/Azure/mmlspark
Kubeflow Fairing (LightGBM on Kubernetes): https://github.com/kubeflow/fairing
ML.NET (.NET/C#-package): https://github.com/dotnet/machinelearning
LightGBM.NET (.NET/C#-package): https://github.com/rca22/LightGBM.Net
Dask-LightGBM (distributed and parallel Python-package): https://github.com/dask/dask-lightgbm
Ruby gem: https://github.com/ankane/lightgbm
- Ask a question on Stack Overflow with the
lightgbmtag, we monitor this for new questions. - Discuss on the LightGBM Gitter.
- Discuss on the LightGBM Slack team.
- Use this invite link to join the team.
- Open bug reports and feature requests (not questions) on GitHub issues.
Check CONTRIBUTING page.
This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.
Guolin Ke, Qi Meng, Thomas Finley, Taifeng Wang, Wei Chen, Weidong Ma, Qiwei Ye, Tie-Yan Liu. "LightGBM: A Highly Efficient Gradient Boosting Decision Tree". Advances in Neural Information Processing Systems 30 (NIPS 2017), pp. 3149-3157.
Qi Meng, Guolin Ke, Taifeng Wang, Wei Chen, Qiwei Ye, Zhi-Ming Ma, Tie-Yan Liu. "A Communication-Efficient Parallel Algorithm for Decision Tree". Advances in Neural Information Processing Systems 29 (NIPS 2016), pp. 1279-1287.
Huan Zhang, Si Si and Cho-Jui Hsieh. "GPU Acceleration for Large-scale Tree Boosting". SysML Conference, 2018.
Note: If you use LightGBM in your GitHub projects, please add lightgbm in the requirements.txt.
This project is licensed under the terms of the MIT license. See LICENSE for additional details.