Filters








1,654 Hits in 5.3 sec

On using prototype reduction schemes to optimize kernel-based nonlinear subspace methods

Sang-Woon Kim, B.John Oommen
2004 Pattern Recognition  
To overcome the limitations of the linear methods, Kernel based Nonlinear Subspace (KNS) methods have been recently proposed in the literature.  ...  Rather than define the matrix K with the whole data set and compute the principal components, we propose that the data be reduced into a smaller representative subset using a Prototype Reduction Scheme  ...  Oja, Subspace Methods of Pattern Recognition, Research Studies Press, 1983.  ... 
doi:10.1016/j.patcog.2003.07.006 fatcat:oisnjcwkbvhehk5dqv6n4q7uxa

On Optimizing Kernel-Based Fisher Discriminant Analysis Using Prototype Reduction Schemes [chapter]

Sang-Woon Kim, B. John Oommen
2006 Lecture Notes in Computer Science  
Rather than invoke the KFDA for the entire data set, we advocate that the data be first reduced into a smaller representative subset using a Prototype Reduction Scheme (PRS), and that dimensionality reduction  ...  Rather than invoke the KFDA for the entire data set, we advocate that the data be first reduced into a smaller representative subset using a Prototype Reduction Scheme (PRS) (explained and briefly surveyed  ...  Rather than define the kernel matrix and compute the principal components using the entire data set, we propose that the size of the data be reduced into a smaller prototype subset using a PRS Since the  ... 
doi:10.1007/11815921_91 fatcat:utxagffjyfc5hgdgnveb7omio4

CANCER MOLECULAR PATTERN DISCOVERY BY SUBSPACE CONSENSUS KERNEL CLASSIFICATION

Xiaoxu Han
2007 Computational Systems Bioinformatics  
We first integrated subspace methods and kernel methods by following our framework of the input space, subspace and kernel space clustering.  ...  The algorithm is a consensus kernel hierarchical clustering (CKHC) method in the subspace generated by the PG-NMF.  ...  Generally, almost all input-space clustering methods can be used in the subspace clustering to cluster the feature data in the subspace.  ... 
doi:10.1142/9781860948732_0010 fatcat:drscwtsesfcm7iudjcfnztqiva

Computational Intelligence-Based Biometric Technologies

D. Zhang, Wangmeng Zuo
2007 IEEE Computational Intelligence Magazine  
CI-based methods, including neural network and fuzzy technologies, have also been extensively investigated for biometric matching.  ...  For those reasons they have been proved to be effective and efficient in biometric feature extraction and biometric matching tasks, sometimes used in combination with traditional methods.  ...  Kernel Dimensionality Reduction Methods Kernel dimensionality reduction (KDR) methods provide an efficient means of handling nonlinear feature extraction problems.  ... 
doi:10.1109/mci.2007.353418 fatcat:aynahy3ttbesfl3qm3u25gcawq

On Using Dimensionality Reduction Schemes to Optimize Dissimilarity-Based Classifiers [chapter]

Sang-Woon Kim, Jian Gao
2008 Lecture Notes in Computer Science  
To avoid these problems, in this paper, we propose an alternative approach where we use all available samples from the training set as prototypes and subsequently apply dimensionality reduction schemes  ...  That is, we prefer not to directly select the representative prototypes from the training samples; rather, we use a dimensionality reduction scheme after computing the dissimilarity matrix with the entire  ...  Here, the dimensionality reduction scheme is used to accommodate some useful information for discrimination and to avoid the problem of finding the optimal prototype selection strategy.  ... 
doi:10.1007/978-3-540-85920-8_38 fatcat:4ior3nbivjdk5hc5lkd3dxzw6e

SAM 2020 Author Index

2020 2020 IEEE 11th Sensor Array and Multichannel Signal Processing Workshop (SAM)  
SS05.2 Adaptive Beamforming Using Frequency Diverse MIMO Radar with Nonlinear Frequency Offset Bartolewska, Julitta R13.4 Distributed Multiarray Noise Reduction With Online Estimation Of Masks  ...  : A Robust Stackelberg Game Perspective R09.6 LPI Performance Optimization Scheme for a Joint Radar-Communications System Wang, Gang SS01.1 A general ESPRIT method for noncircularity-based incoherently  ... 
doi:10.1109/sam48682.2020.9104397 fatcat:cfp5gsikrzabhhcnkalahjkxze

Density-matrix based Extended Lagrangian Born-Oppenheimer Molecular Dynamics [article]

Anders M. N. Niklasson
2020 arXiv   pre-print
The integration scheme is based on a tunable, low-rank approximation of a fourth-order kernel, K, that determines the metric tensor, T≡ K^T K, used in the extended harmonic oscillator of the Lagrangian  ...  The formulation and algorithms provide a general guide to implement extended Lagrangian Born-Oppenheimer molecular dynamics for quantum chemistry, density functional theory, and semiempirical methods using  ...  The orthogonalized rank-m Krylov subspace approximation of the kernel K in the integration of the electronic degrees of freedom, Eq. 30-32, based on Ref. [37] is used.  ... 
arXiv:2003.09050v4 fatcat:srwapwcncbhyjb464q37d5ko7a

Gabor-Based Kernel Partial-Least-Squares Discrimination Features for Face Recognition

Vitomir Štruc, Nikola Pavešić
2009 Informatica  
For optimal performance, the KPLSD-based transformation is implemented using the recently proposed fractional-power-polynomial models.  ...  The proposed feature-extraction method, called the Gabor-based kernel partial-least-squares discrimination (GKPLSD), is performed in two consecutive steps.  ...  ., 2003) , PCA is an optimal compression scheme that minimizes the mean squared error between the original training data and its reconstruction for any given level of compression.  ... 
doi:10.15388/informatica.2009.240 fatcat:4grgyhh4lvcbrlp7uiarquzqhm

Automated estimation of tumor probability in prostate magnetic resonance spectroscopic imaging: Pattern recognition vs quantification

B. Michael Kelm, Bjoern H. Menze, Christian M. Zechmann, Klaus T. Baudendistel, Fred A. Hamprecht
2006 Magnetic Resonance in Medicine  
Furthermore, linear methods based on magnitude spectra easily achieve equal performance and also allow for biochemical interpretation in combination with subspace methods.  ...  Nonlinear methods operating directly on magnitude spectra achieve the best results but are less transparent than the linear methods.  ...  Optimal subspaces along with the most important spectral patterns are automatically determined based on in vivo data.  ... 
doi:10.1002/mrm.21112 pmid:17191229 fatcat:mehbvdytsrfsjkjtzcogroylke

The Kernel Common Vector Method: A Novel Nonlinear Subspace Classifier for Pattern Recognition

H.. Cevikalp, M.. Neamtu, A.. Barkana
2007 IEEE Transactions on Systems Man and Cybernetics Part B (Cybernetics)  
methods as well as the kernel-based nonlinear subspace method.  ...  Index Terms-Common vector (CV), kernel-based subspace method, pattern recognition, subspace classifier.  ...  The kernel-based nonlinear subspace method, which is called the kernel CLAFIC, was developed to overcome this limitation [16] , [17] .  ... 
doi:10.1109/tsmcb.2007.896011 pmid:17702291 fatcat:axbwadxuhnegxgi2ebxge7t4ne

Data-Derived Analysis and Inference for an Industrial Deethanizer

Francesco Corona, Michela Mulas, Roberto Baratti, Jose A. Romagnoli
2012 Industrial & Engineering Chemistry Research  
The SOM and the MLP are classic methods for nonlinear dimensionality reduction and nonlinear function estimation widely adopted in process systems engineering; here, the effectiveness of these data derived  ...  The discussed methods are used in visualizing process measurements, extracting operational information and designing an estimation model.  ...  The approach is based on a classical machine learning method for dimensionality reduction and quantization, the Self-Organizing Map, SOM (Kohonen, 2001) .  ... 
doi:10.1021/ie202854b fatcat:gfskdybvkfhl3jtqy4gup47g24

Manifold Based Local Classifiers: Linear and Nonlinear Approaches

Hakan Cevikalp, Diane Larlus, Marian Neamtu, Bill Triggs, Frederic Jurie
2008 Journal of Signal Processing Systems  
This procedure allows us to use a wide variety of distance functions in the process, while computing distances between the query sample and the nonlinear manifolds remains straightforward owing to the  ...  In this paper we reformulate HKNN in terms of subspaces, and propose a variant, the Local Discriminative Common Vector (LDCV) method, that is more suitable for classification tasks where the classes have  ...  Then, based on the subspace formulation, the HKNN method has been extended to the nonlinear case using the kernel trick. However, the nonlinearization of the method is not trivial.  ... 
doi:10.1007/s11265-008-0313-4 fatcat:dthcsdj2o5fz3jxqfysguna7gi

Data derived analysis and inference for an industrial deethanizer

Francesco Corona, Michela Mulas, Roberto Baratti, Jose A. Romagnoli
2009 IFAC Proceedings Volumes  
The SOM and the MLP are classic methods for nonlinear dimensionality reduction and nonlinear function estimation widely adopted in process systems engineering; here, the effectiveness of these data derived  ...  The discussed methods are used in visualizing process measurements, extracting operational information and designing an estimation model.  ...  The approach is based on a classical machine learning method for dimensionality reduction and quantization, the Self-Organizing Map, SOM (Kohonen, 2001) .  ... 
doi:10.3182/20090712-4-tr-2008.00111 fatcat:rcrk3zctpzem5npbaj5hhyp4ai

Super-sparse Learning in Similarity Spaces [article]

Ambra Demontis, Marco Melis, Battista Biggio, Giorgio Fumera, Fabio Roli
2017 arXiv   pre-print
We overcome this limitation by jointly learning the classification function along with an optimal set of virtual prototypes, whose number can be either fixed a priori or optimized according to application-specific  ...  Current reduction approaches select a small subset of representative prototypes in the space induced by the similarity measure, and then separately train the classification function on the reduced subset  ...  Our method can also be used to reduce the number of prototypes used by kernel-based or prototype-based classifiers, like the SVM, by setting the target variables y to the values of the discriminant function  ... 
arXiv:1712.06131v1 fatcat:vgr7a74wnveirneer2dbeuxhki

Missing Image Data Reconstruction Based on Adaptive Inverse Projection via Sparse Representation

Takahiro Ogawa, Miki Haseyama
2011 IEEE transactions on multimedia  
Furthermore, in this approach, the proposed method monitors errors caused by the derived inverse projection, and the lowdimensional subspaces optimal for target textures are adaptively selected.  ...  The proposed method also introduces some schemes for color processing into the calculation of subspaces on the basis of sparse representation and attempts to avoid spurious color caused in the reconstruction  ...  Thus, several missing texture reconstruction methods that utilize projection schemes onto nonlinear subspaces obtained by kernel PCA and CCA have been proposed [24] , [25] .  ... 
doi:10.1109/tmm.2011.2161760 fatcat:7kmuy77pyzc25mniwv333ipcke
« Previous Showing results 1 — 15 out of 1,654 results