Duet is a decision tree ensemble method based multiclass classification framework that offers a more efficient resource usage while preserving and even improving the classification accuracy in comparison to standard monolithic models.
Duet is based on a small bagging ensemble model and a booting model.
The current implementation of Duet is based on Random Forest and XGBoost.
More details about the Duet can be found in the following paper:
"Efficient Multiclass Classification with Duet" [EuroMLSys '22]
https://dl.acm.org/doi/abs/10.1145/3517207.3526970
https://euromlsys.eu/pdf/euromlsys22-final4.pdf
Duet scikit classifier
Basic classification example by Duet
Basic grid search example with Duet
numpy
pandas
skleran
xgboost
or alternatively, run:
$ pip3 install -r requirements.txt
The efficient-multiclass-classification project team welcomes contributions from the community. Before you start working with efficient-multiclass-classification, please read our Developer Certificate of Origin. All contributions to this repository must be signed as described on that page. Your signature certifies that you wrote the patch or have the right to pass it on as an open-source patch. For more detailed information, refer to CONTRIBUTING.md.
BSD-3 License
For more information, support and advanced examples contact:
Yaniv Ben-Itzhak, ybenitzhak@vmware.com
Shay Vargaftik, shayv@vmware.com