| Autor: |
Biliri Evmorfia, Lampathaki Fenareti, Mandilaras George, Prieto-Roig Ausias, Calabresi Mattia, Branco Rui, Gkolemis Vasileios |
| Rok vydání: |
2023 |
| Předmět: |
|
| DOI: |
10.5281/zenodo.8010283 |
| Popis: |
Lack of understanding for machine learningmodels’ inner workings by business users inevitably leads to lack of trust, particularly in critical operations. Recent advancements in artificial intelligence include the development of explainability methods and tools for machine learning and deep learning models. These explainable AI (XAI) techniques can significantly reduce the black box effect that often hinders direct inclusion and automated integration of ML outputs in decision making. The manufacturing sector is gradually increasing its adoption of AI-enabled systems, a process accelerated in the context of the Fourth Industrial Revolution, strengthening the need for explainability and robust XAI pipelines. This paper presents a framework of processes and tools designed and developed to bring the benefits of explainability in AI-enabled decision making in the manufacturing domain, providing the mechanisms to create production ready, trustful machine learning processes that foster collaboration among stakeholders coming from both business and technical backgrounds. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
