Towards Reproducibility in Predictive Process Mining: SPICE -- A Deep Learning Library
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866911327338889216 |
|---|---|
| author | Stritzel, Oliver Hühnerbein, Nick Rauch, Simon Zarate, Itzel Fleischmann, Lukas Buck, Moike Lischka, Attila Frey, Christian |
| author_facet | Stritzel, Oliver Hühnerbein, Nick Rauch, Simon Zarate, Itzel Fleischmann, Lukas Buck, Moike Lischka, Attila Frey, Christian |
| contents | In recent years, Predictive Process Mining (PPM) techniques based on artificial neural networks have evolved as a method for monitoring the future behavior of unfolding business processes and predicting Key Performance Indicators (KPIs). However, many PPM approaches often lack reproducibility, transparency in decision making, usability for incorporating novel datasets and benchmarking, making comparisons among different implementations very difficult. In this paper, we propose SPICE, a Python framework that reimplements three popular, existing baseline deep-learning-based methods for PPM in PyTorch, while designing a common base framework with rigorous configurability to enable reproducible and robust comparison of past and future modelling approaches. We compare SPICE to original reported metrics and with fair metrics on 11 datasets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_16715 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Towards Reproducibility in Predictive Process Mining: SPICE -- A Deep Learning Library Stritzel, Oliver Hühnerbein, Nick Rauch, Simon Zarate, Itzel Fleischmann, Lukas Buck, Moike Lischka, Attila Frey, Christian Machine Learning Artificial Intelligence In recent years, Predictive Process Mining (PPM) techniques based on artificial neural networks have evolved as a method for monitoring the future behavior of unfolding business processes and predicting Key Performance Indicators (KPIs). However, many PPM approaches often lack reproducibility, transparency in decision making, usability for incorporating novel datasets and benchmarking, making comparisons among different implementations very difficult. In this paper, we propose SPICE, a Python framework that reimplements three popular, existing baseline deep-learning-based methods for PPM in PyTorch, while designing a common base framework with rigorous configurability to enable reproducible and robust comparison of past and future modelling approaches. We compare SPICE to original reported metrics and with fair metrics on 11 datasets. |
| title | Towards Reproducibility in Predictive Process Mining: SPICE -- A Deep Learning Library |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2512.16715 |