XGBoost Algorithm
XGBoost Algorithm stands for eXtreme Gradient Boosting. XGBoost uses ensemble learning, which is also called boosting. The results of multiple models are grouped together to produce a better fit to the training data. Each decision tree model is added using the prediction errors of previous models to improve the fit to the training data.
XGBoost Algorithm can be used four types of problems:
- Regression
- Binary classification
- Multi-class classification
- Ranking
XGBoost Algorithm is an open source library available for multiple programming languages. Because of this specific releases are available as a SageMaker built-in algorithm and it can be used as a built-in algorithm or framework.
- AWS docs: https://docs.aws.amazon.com/sagemaker/latest/dg/xgboost.html
- XGBoost website: https://xgboost.ai/
- XGBoost explanation: https://xgboost.readthedocs.io/en/latest/tutorials/model.html
- Wikipedia: https://en.wikipedia.org/wiki/XGBoost
- https://machinelearningmastery.com/extreme-gradient-boosting-ensemble-in-python/
Attributes
| Problem attribute | Description |
| Data types and format | Tabular |
| Learning paradigm or domain | Supervised Learning |
| Problem type | Binary/multi-class classification, Regression |
| Use case examples | Predict a numeric/continuous value; Predict if an item belongs to a category |
Training
Training data for the XGBoost Algorithm is CSV or libsvm formats.
Model artifacts and inference
| Description | Artifacts |
| Learning paradigm | Supervised Learning |
| Request format | Tabular data in CSV or libsvm format |
Processing environment
- Training: single CPU or GPU (memory bound)
- Inference: not known
Video
Amazon SageMaker’s Built-in Algorithm Webinar Series: XGBoost
- 1.01.01 minutes
- 0-25.31 Introduction to XGBoost
- 25.32+ demo
Credits
Rugby photo by Philippa Rose-Tite on Unsplash
