Fault Detection Using Nonlinear Low-Dimensional Representation of Sensor Data
This work addresses fault detection for IoT-based monitoring systems, but it is incremental as it applies existing methods to a specific domain.
The paper tackled fault detection in critical equipment by applying nonlinear dimension reduction techniques like t-SNE and KPCA to sensor data, resulting in improved interpretability and suitability for edge processing in IoT applications.
Sensor data analysis plays a key role in health assessment of critical equipment. Such data are multivariate and exhibit nonlinear relationships. This paper describes how one can exploit nonlinear dimension reduction techniques, such as the t-distributed stochastic neighbor embedding (t-SNE) and kernel principal component analysis (KPCA) for fault detection. We show that using anomaly detection with low dimensional representations provides better interpretability and is conducive to edge processing in IoT applications.