Concept Drift Detection with Variable Interaction Networks
This work addresses predictive maintenance for industrial systems, but it appears incremental as it builds on existing sliding window methods for drift detection.
The paper tackles the problem of detecting malfunctions in production plants by monitoring changes in variable interactions, presenting a sliding window algorithm that successfully identifies concept drift in a synthetic dynamical system.
The current development of today's production industry towards seamless sensor-based monitoring is paving the way for concepts such as Predictive Maintenance. By this means, the condition of plants and products in future production lines will be continuously analyzed with the objective to predict any kind of breakdown and trigger preventing actions proactively. Such ambitious predictions are commonly performed with support of machine learning algorithms. In this work, we utilize these algorithms to model complex systems, such as production plants, by focusing on their variable interactions. The core of this contribution is a sliding window based algorithm, designed to detect changes of the identified interactions, which might indicate beginning malfunctions in the context of a monitored production plant. Besides a detailed description of the algorithm, we present results from experiments with a synthetic dynamical system, simulating stable and drifting system behavior.