This repository contains supplementary material for the article "Prescriptive Process Monitoring for Cost-Aware Cycle Time Reduction" by Zahra Dasht Bozorgi, Irene Teinemaa, Marlon Dumas, Marcello La Rosa, and Artem Polyvyanyy.

Reference

If you use the code from this repository, please cite the original paper:

@article{bozorgi2021prescriptive,
  title={Prescriptive Process Monitoring for Cost-Aware Cycle Time Reduction},
  author={Bozorgi, Zahra Dasht and Teinemaa, Irene and Dumas, Marlon and La Rosa, Marcello},
  journal={arXiv preprint arXiv:2105.07111},
  year={2021}
}

Dependencies

• python 3.7
• NumPy
• pandas
• scikit-learn
• EconML (for causal models construction)

Preprocessing

For ease of use, the preprocessed datasets for BPIC17 and BPIC19 can be found here and here respectively. These datasets include the cycle time, treatment indicator and the potentially confounding features as discussed in the paper.

Experiments

The results of our experiments can be reproduced by running all the cells of bpi2017_experiments.ipynb and bpi2019_experiments.ipynb.