The Case for Explainability of Real-Time Systems and their Analyses
Abstract
Practitioners do not trust output of analysis because they don't know how the analysis works. They don't know if analysis is correctly implemented. They don't know if the assumptions that the analysis rely on applies to the system at hand; answer the question whether assumptions apply requires, in turn, an analysis or understanding of analysis (which practitioners don't have). They don't know if they have entered the model correctly.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 31, 2022
- Accession Number
- AD1183614
Entities
People
- Björn Andersson
Organizations
- Carnegie Mellon University