A Software Interface for Machine Learning Model Applications

Fuente: Zenodo
Salvato in:
Dettagli Bibliografici
Autori principali: Konomos, Antonios, Chountasis, Spiros
Natura: Recurso digital
Pubblicazione: Zenodo 2026
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866901913714294784
author Konomos, Antonios
Chountasis, Spiros
author_facet Konomos, Antonios
Chountasis, Spiros
contents <p><span lang="EN">This paper presents an innovative software interface for the utilization of widely used Machine Learning (ML) algorithms in a unified Python/R programming environment. A novel software model, the Rbox+, is proposed to execute ML algorithms by jointly leveraging the capabilities of the Python and R programming languages. Furthermore, more comprehensive and specialized architecture is made available for integrating ML into enterprise information systems. Unlike conventional ML Application Programming Interfaces (APIs) or isolated Enterprise Resource Planning (ERP) analytics tools, the Rbox+ enables transparent, language-independent execution and validation of ML models while exposing the underlying source code. The proposed approach supports practical applications in enterprise analytics, reproducible research, and enhancing interoperability between ERP systems, analytics platforms, and statistical programming environments. The proposed API has been tested and evaluated using a publicly available benchmark dataset for regression analysis, applying multiple ML models and comparative performance metrics. The obtained results demonstrate improved computational efficiency and scalability.</span></p>
format Recurso digital
id zenodo_https___doi_org_10_5281_zenodo_18831487
institution Zenodo
language
publishDate 2026
publisher Zenodo
record_format zenodo
spellingShingle A Software Interface for Machine Learning Model Applications
Konomos, Antonios
Chountasis, Spiros
<p><span lang="EN">This paper presents an innovative software interface for the utilization of widely used Machine Learning (ML) algorithms in a unified Python/R programming environment. A novel software model, the Rbox+, is proposed to execute ML algorithms by jointly leveraging the capabilities of the Python and R programming languages. Furthermore, more comprehensive and specialized architecture is made available for integrating ML into enterprise information systems. Unlike conventional ML Application Programming Interfaces (APIs) or isolated Enterprise Resource Planning (ERP) analytics tools, the Rbox+ enables transparent, language-independent execution and validation of ML models while exposing the underlying source code. The proposed approach supports practical applications in enterprise analytics, reproducible research, and enhancing interoperability between ERP systems, analytics platforms, and statistical programming environments. The proposed API has been tested and evaluated using a publicly available benchmark dataset for regression analysis, applying multiple ML models and comparative performance metrics. The obtained results demonstrate improved computational efficiency and scalability.</span></p>
title A Software Interface for Machine Learning Model Applications
url https://doi.org/10.5281/zenodo.18831487