Knowledge Compilation Techniques for Model-Based Diagnosis of Complex Active Systems - Machine Learning and Knowledge Extraction Access content directly
Conference Papers Year : 2018

Knowledge Compilation Techniques for Model-Based Diagnosis of Complex Active Systems

Abstract

According to complexity science, the essence of a complex system is the emergence of unpredictable behavior from interaction among components. Loosely inspired by this idea, a diagnosis technique of a class of discrete-event systems, called complex active systems, is presented. A complex active system is a hierarchical graph, where each node is a network of communicating automata, called an active unit. Specific interaction patterns among automata within an active unit give rise to the occurrence of emergent events, which may affect the behavior of superior active units. This results in the stratification of the behavior of the complex active system, where each different stratum corresponds to a different abstraction level of the emergent behavior. As such, emergence is a peculiar property of a complex active system. To speed up the diagnosis task, model-based knowledge is compiled offline and exploited online by the diagnosis engine. The technique is sound and complete.
Fichier principal
Vignette du fichier
472936_1_En_4_Chapter.pdf (744 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-02060059 , version 1 (07-03-2019)

Licence

Attribution

Identifiers

Cite

Gianfranco Lamperti, Marina Zanella, Xiangfu Zhao. Knowledge Compilation Techniques for Model-Based Diagnosis of Complex Active Systems. 2nd International Cross-Domain Conference for Machine Learning and Knowledge Extraction (CD-MAKE), Aug 2018, Hamburg, Germany. pp.43-64, ⟨10.1007/978-3-319-99740-7_4⟩. ⟨hal-02060059⟩
57 View
53 Download

Altmetric

Share

Gmail Facebook X LinkedIn More