An orthogonally filtered tree classifier based on nonlinear kernel-based optimal representation of data

The early detection and reliable diagnosis of a fault is crucial in an on-going operation of processes. They provide early warning for a fault and identification of its assignable cause. This paper proposes a classification tree-based diagnosis scheme combined with nonlinear kernel discriminant analysis. The nonlinear kernel-based dimension reduction for the discrimination of various classes of data is performed to determine nonlinear decision boundaries. The use of the nonlinear kernel method in a classification tree is to reduce the dimension of data and to provide its lower-dimensional representation suitable for separating different classes. We also present the use of orthogonal filter as a preprocessing step. An orthogonal filter-based preprocessing is performed to remove unwanted variation of data for enhancing discrimination power and classification performance. The performance of the proposed method is demonstrated using simulation data and compared with other methods. The classification results showed that the proposed tree-based method outperforms traditional PCA-based method.

[1]  Ali Cinar,et al.  Statistical process monitoring and disturbance diagnosis in multivariable continuous processes , 1996 .

[2]  Leo H. Chiang,et al.  Fault diagnosis in chemical processes using Fisher discriminant analysis, discriminant partial least squares, and principal component analysis , 2000 .

[3]  E. F. Vogel,et al.  A plant-wide industrial process control problem , 1993 .

[4]  Bernhard Schölkopf,et al.  Nonlinear Component Analysis as a Kernel Eigenvalue Problem , 1998, Neural Computation.

[5]  P. Bishnoi,et al.  Fault diagnosis of multivariate systems using pattern recognition and multisensor data analysis technique , 2001 .

[6]  John F. MacGregor,et al.  Multivariate SPC charts for monitoring batch processes , 1995 .

[7]  G. Baudat,et al.  Generalized Discriminant Analysis Using a Kernel Approach , 2000, Neural Computation.

[8]  Roman Rosipal,et al.  Kernel Partial Least Squares Regression in Reproducing Kernel Hilbert Space , 2002, J. Mach. Learn. Res..

[9]  S. Joe Qin,et al.  Statistical process monitoring: basics and beyond , 2003 .

[10]  Age K. Smilde,et al.  Direct orthogonal signal correction , 2001 .

[11]  Theodora Kourti,et al.  Analysis, monitoring and fault diagnosis of batch processes using multiblock and multiway PLS , 1995 .

[12]  E. K. Kemsley,et al.  Discriminant analysis of high-dimensional data: a comparison of principal components analysis and partial least squares data reduction methods , 1996 .

[13]  Michael I. Jordan,et al.  Kernel independent component analysis , 2003 .

[14]  Manabu Kano,et al.  Monitoring independent components for fault detection , 2003 .

[15]  S. Wold,et al.  Orthogonal signal correction of near-infrared spectra , 1998 .

[16]  Christos Georgakis,et al.  Disturbance detection and isolation by dynamic principal component analysis , 1995 .

[17]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[18]  Thomas J. McAvoy,et al.  Base control for the Tennessee Eastman problem , 1994 .

[19]  D. Kell,et al.  Classification of pyrolysis mass spectra by fuzzy multivariate rule induction-comparison with regression, K-nearest neighbour, neural and decision-tree methods , 1997 .

[20]  Rasmus Bro,et al.  Orthogonal signal correction, wavelet analysis, and multivariate calibration of complicated process fluorescence data , 2000 .

[21]  Gunnar Rätsch,et al.  An introduction to kernel-based learning algorithms , 2001, IEEE Trans. Neural Networks.

[22]  Gerhard Tutz,et al.  A CART-based approach to discover emerging patterns in microarray data , 2003, Bioinform..