BBF [1] is an fine-tuning approach that bases on boosting [2] and Bagging [3]. It is designed to be efficient with the following advantages:
- Support of fine-tuning an initial model for classification.
- Support of fine-tuning an initial model for regression.
Our primary documentation is at https://bbf.readthedocs.io and is generated from this repository. If you are new to BBF, follow the installation instructions on that site. The preferred way to install BBF is via pip from Pypi.
Next you may want to read:
- APIs & Parameters is an exhaustive list of customization you can make.
- Parameters Tuning is an exhaustive list of customization you can make.
- Examples showing command line usage of common tasks.
[1] C. Zhao, R. Peng and D. Wu, “Bagging and Boosting Fine-tuning for Ensemble Learning,” IEEE Trans. on Artificial Intelligence, Early Access, 2023, DOI: 10.1109/TAI.2023.3296685.
[2] J. H. Friedman, “Greedy function approximation: A gradient boosting machine,” Annals of Statistics, vol. 29, no. 5, pp. 1189–1232, 2001.
[3] L. Breiman, “Bagging predictors,” Machine Learning, vol. 24, no. 2, pp. 123–140, 1996.