Emergent behaviour in a system of industrial plants detected via manifold learning

Abstract

The efficiency behaviour of an industrial plant, part of a huge international structure of plants, is modelled as an emergent phenomenon in a complex adaptive system. The study is based on real in-service data obtained from an industrial production line monitoring system. Models of complex adaptive systems and some modern manifold learning methods are introduced in a unified formalism. The emergent behaviour is efficiently described in this setup.

Keywords

• spectral analysis
• spectral clustering
• manifold learning
• emergent behaviour
• nonlocal features