This is a Python implementation of the Algorithm DJEnsemble in the following paper: DJEnsemble: a Cost-Based Selection and Allocation of a Disjoint Ensemble of Spatio-temporal Models by Rafael Pereira, Yania Souto, Anderson Chaves, Rocio Zorilla, Brian Tsan, Florin Rusu, Eduardo Ogasawara, Artur Ziviani, Fabio Porto. SSDBM 2021
The full version of the paper can be accessed at arXiv.
Disclaimer: this is a unified Python implementation of the original work presented at SSDBM,
- dtw==1.4.0
- dtw_python==1.1.12
- gldpy==0.2
- h5py==3.6.0
- matplotlib==3.1.2
- numpy==1.17.4
- pandas==0.25.3
- scikit_learn==1.1.2
- tensorflow==2.7.0
- xarray==0.20.2
Dependencies can be installed using the following command:
pip install -r requirements.txtThe projet should be run from main.py, informing the configuration file. The example selects the best ensemble for a set of three models trained on the alerta-rio dataset:
python main.py queries/query-alerta-rio.config