Digital Library

Amllibrary: an automl approach for performance prediction

• Bruno Guindani
• Marco Lattuada
• Danilo Ardagna

(2023). ECMS 2023, 37th Proceedings
Edited by: Enrico Vicario, Romeo Bandinelli, Virginia Fani, Michele Mastroianni, European Council for Modelling and Simulation.
DOI: http://doi.org/10.7148/2023
ISSN: 2522-2422 (ONLINE)
ISSN: 2522-2414 (PRINT)
ISSN: 2522-2430 (CD-ROM)
ISBN: 978-3-937436-80-7
ISBN: 978-3-937436-79-1 (CD) Communications of the ECMS Volume 37, Issue 1, June 2023, Florence, Italy June 20th – June 23rd, 2023

https://doi.org/10.7148/2023-0215

Bruno guindani, Marco lattuada, Danilo ardagna (2023). AMLLibrary: An AutoML Approach For Performance Prediction, ECMS 2023, Proceedings Edited by: Enrico Vicario, Romeo Bandinelli, Virginia Fani, Michele Mastroianni, European Council for Modelling and Simulation. doi:10.7148/2023-0215

aMLLibrary is an open-source, high-level Python package that allows the parallel building of multiple Machine Learning (ML) regression models. It is focused on performance modeling and supports several methods for feature engineering/selection and hyperparameter tuning. The library implements fault tolerance mechanisms to recover from system crashes, and only a simple declarative text file is required to launch a full experimental campaign for all required models. Its modular structure allows users to implement their own plugins and model-building wrappers and easily add them to the library. We test aMLLibrary on building the performance models of neural networks and image processing applications, with the best model produced often having less than 20% prediction error.