Normalized Cuts and Image Segmentation
IEEE Transactions on Pattern Analysis and Machine Intelligence
Laplacian Eigenmaps for dimensionality reduction and data representation
Neural Computation
Spectral clustering and its use in bioinformatics
Journal of Computational and Applied Mathematics
| Hi-index | 0.00 |
The statistical process control of certain complex production processes or products rely on the monitoring of profiles. In this work, we consider a spectral-based method to distinguish between outlier and nonoutlier observations to establish profile parameters. The proposed method is based on the spectral theory of the Laplacian of a graph, more specifically, a Laplacian eigenmap. The graph is constructed by considering each profile as a node, and by adjusting edge weights on the graph to reflect the resemblance between profiles. Laplacian eigenmaps have been used to represent data, to reduce data dimension, and for clustering. The proposed method can be used for complex profiles because it does not require the function between the dependent and independent variables. We apply the proposed method to an artificially generated data set from a nonlinear profile and to a wood board production problem. The results are compared with existing methods in the literature and show that the proposed method performs satisfactorily. Copyright © 2012 John Wiley & Sons, Ltd.