Dynamic SHAP: Time resolved explainable AI for batch processes

Computers & Chemical Engineering 2026 Open access

Read full text
Authors: Luca Brunaccioni, Francesco Fusco, Mattia Vallerio
DOI: 10.1016/j.compchemeng.2026.109717
Cited by: 1
References: 37

Abstract

• Introduces Dynamic SHAP (D-SHAP) for time-resolved interpretability of batch-process ML models. • D-SHAP reveals when and which variables most influence product quality, closing the accuracy–explainability gap. • Methodology validated on one industrial and two synthetic datasets, confirming broad applicability.

Fields of Study & Subjects

Explainable Artificial Intelligence (XAI) Machine Learning in Materials Science Generative Adversarial Networks and Image Synthesis

Publication details

Journal / source
Computers & Chemical Engineering
Publisher
Elsevier BV
Volume
213
Pages
109717–109717
Publication date
29 May 2026
Language
EN