VERONA: predictiVE pRocess mOnitoring beNchmArk

The VERONA library is a Python tool for faster and more user-friendly development, evaluation, and comparison of predictive process monitoring (PPM) approaches.

The main features of VERONA are:

• Allows downloading an extensive collection of real process event logs used by the process mining community.

• Provides functions to cover data preprocessing, partitioning, prefix and target extraction, model evaluation and result visualization from the PPM workflow in a clear and transparent manner for the user.

• Implements the reference benchmark of Rama-Maneiro et al. [1] to fairly compare deep learning-based PPM approaches.

• Ports the statistical tests of the R-implemented scmamp [2] to Python, allowing statistical comparison of different machine learning approaches.

Note

This project is under active development.

Contents

• Overview
  • Introduction to Predictive Process Monitoring
  • The PPM workflow
  • Benchmarking deep learning-based approaches
  • The importance of VERONA library
• Installation
  • PyPI
  • Manual compilation
• Software architecture
• API Reference
  • verona.data package
  • verona.evaluation package
  • verona.visualization package
• Cite the paper

References

[1]

Efrén Rama-Maneiro, Juan C. Vidal, and Manuel Lama. Deep learning for predictive business process monitoring: review and benchmark. IEEE Transactions on Services Computing, 16(1):739–756, 2023. doi:10.1109/TSC.2021.3139807.

[2]

Borja Calvo and Guzmán Santafé. Scmamp: statistical comparison of multiple algorithms in multiple problems. R J., 8(1):248, 2016. URL: https://doi.org/10.32614/rj-2016-017, doi:10.32614/RJ-2016-017.

Indices and tables

• Index

• Module Index

• Search Page