Abstract: High-dimensional data streams are becoming increasingly ubiquitous in industrial systems. Efficient detection of system faults from these data can ensure the reliability and safety of the ...
Abstract: This letter evaluates the performance of a nonsupervised clustering method for identifying abrupt changepoints in streaming sensor data. The proposed method utilizes the Jenks natural breaks ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results