A Model-Based Approach to Experiment-Driven Evolution of ML Workflows

Petr Hnětynka, Tomáš Bureš, Ilias Gerostathopoulos, Milad Abdullah, Keerthiga Rajenthiram

2025

Abstract

Machine Learning (ML) has advanced significantly, yet the development of ML workflows still relies heavily on expert intuition, limiting standardization. MLOps integrates ML workflows for reliability, while AutoML automates tasks like hyperparameter tuning. However, these approaches often overlook the iterative and experimental nature of the development of ML workflows. Within the ongoing ExtremeXP project (Horizon Europe), we propose an experiment-driven approach where systematic experimentation becomes central to ML workflow evolution. The framework created within the project supports transparent, reproducible, and adaptive experimentation through a formal metamodel and related domain-specific language. Key principles include traceable experiments for transparency, empowered decision-making for data scientists, and adaptive evolution through continuous feedback. In this paper, we present the framework from the model-based approach perspective. We discuss the lessons learned from the use of the metamodel-centric approach within the project—especially with use-case partners without prior modeling expertise.

Paper Citation

in Harvard Style

Hnětynka P., Bureš T., Gerostathopoulos I., Abdullah M. and Rajenthiram K. (2025). A Model-Based Approach to Experiment-Driven Evolution of ML Workflows. In Proceedings of the 13th International Conference on Model-Based Software and Systems Engineering - Volume 1: MODELSWARD; ISBN 978-989-758-729-0, SciTePress, pages 354-362. DOI: 10.5220/0013380500003896

in Bibtex Style

@conference{modelsward25,
author={Petr Hnětynka and Tomáš Bureš and Ilias Gerostathopoulos and Milad Abdullah and Keerthiga Rajenthiram},
title={A Model-Based Approach to Experiment-Driven Evolution of ML Workflows},
booktitle={Proceedings of the 13th International Conference on Model-Based Software and Systems Engineering - Volume 1: MODELSWARD},
year={2025},
pages={354-362},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013380500003896},
isbn={978-989-758-729-0},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 13th International Conference on Model-Based Software and Systems Engineering - Volume 1: MODELSWARD
TI - A Model-Based Approach to Experiment-Driven Evolution of ML Workflows
SN - 978-989-758-729-0
AU - Hnětynka P.
AU - Bureš T.
AU - Gerostathopoulos I.
AU - Abdullah M.
AU - Rajenthiram K.
PY - 2025
SP - 354
EP - 362
DO - 10.5220/0013380500003896
PB - SciTePress