/pdm

Predictive deviation monitoring (pdm)

Primary LanguagePython

Predictive deviation monitoring (pdm)

General

PDM is an outcome-oriented predictive business process monitoring method for predictive deviation monitoring using conformance checking and deep learning.

Paper

If you use our code or fragments of it, please cite our paper:

@proceedings{weinzierl2021pdm,
    title={Predictive Business Process Deviation Monitoring},
    author={Sven Weinzierl and Sebastian Dunzer and Johannes Tenschert and Sandra Zilker and Martin Matzner},
    booktitle={Proceedings of the 29th European Conference on Information Systems (ECIS2021)},
    year={2021}
}

Setup

  1. Install Miniconda (https://docs.conda.io/en/latest/miniconda.html)
  2. After setting up miniconda you can make use of the conda command in your command line (e.g. CMD or Bash)
  3. To quickly install the pdm package, run pip install -e . inside the root directory.
  4. To install required packages run pip install -r requirements.txt inside the root directory.
  5. Set parameters in config.py
  6. Train and test the Bi-LSTM models for the next activity prediction by executing runner.py