Explainability for Fault Detection System in Chemical Processes
Georgios Gravanis, Dimitrios Kyriakou, Spyros Voutetakis, Simira Papadopoulou, Konstantinos Diamantaras · 2026
Read Explainability for Fault Detection System in Chemical Processes by Georgios Gravanis, Dimitrios Kyriakou, Spyros Voutetakis, Simira Papadopoulou, Konstantinos Diamantaras (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-16341.txt.
Description
Georgios Gravanis leads a team of four coauthors comparing two explainability methods for fault detection in chemical processes. The team applies Integrated Gradients and SHapley Additive exPlanations to explain the decisions of a highly accurate Long Short-Term Memory classifier. That classifier is trained to spot faults in the Tennessee Eastman Process, a benchmark nonlinear chemical process.
The explainability methods help pinpoint the subsystem where a fault occurred. In most cases both methods flag the same features as most important, while in some cases SHAP looks more informative and closer to the root cause. Since both methods are model-agnostic, the approach is not tied to this one process and can be carried over to similar problems.
Catalog record
| CREATOR | Georgios Gravanis, Dimitrios Kyriakou, Spyros Voutetakis, Simira Papadopoulou, Konstantinos Diamantaras |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 23 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-16341