Kernel PCA for novelty detection

Authors:
Heiko Hoffmann
Affiliations:
Max Planck Institute for Human Cognitive and Brain Sciences, Amalienstr. 33, 80799 Munich, Germany
Venue:
Pattern Recognition
Year:
2007

Citing 13
Cited 32

Numerical recipes in C (2nd ed.): the art of scientific computing

Numerical recipes in C (2nd ed.): the art of scientific computing
Principal component neural networks: theory and applications

Principal component neural networks: theory and applications
Nonlinear component analysis as a kernel eigenvalue problem

Neural Computation
Mixtures of probabilistic principal component analyzers

Neural Computation
Support vector domain description

Pattern Recognition Letters - Special issue on pattern recognition in practice VI
Kernel PCA and de-noising in feature spaces

Proceedings of the 1998 conference on Advances in neural information processing systems II
Robust De-noising by Kernel PCA

ICANN '02 Proceedings of the International Conference on Artificial Neural Networks
Kernel Whitening for One-Class Classification

SVM '02 Proceedings of the First International Workshop on Pattern Recognition with Support Vector Machines
Fast Approximation of Support Vector Kernel Expansions, and an Interpretation of Clustering as Approximation in Feature Spaces

Mustererkennung 1998, 20. DAGM-Symposium
Novelty detection: a review—part 1: statistical approaches

Signal Processing
Novelty detection: a review—part 2: neural network based approaches

Signal Processing
Estimating the Support of a High-Dimensional Distribution

Neural Computation
Approaches for automated detection and classification of masses in mammograms

Pattern Recognition

Supporting diagnosis of attention-deficit hyperactive disorder with novelty detection

Artificial Intelligence in Medicine
A boundary method for outlier detection based on support vector domain description

Pattern Recognition
Automatic target defect identification for TFT-LCD array process inspection using kernel FCM-based fuzzy SVDD ensemble

Expert Systems with Applications: An International Journal
Stacking for Ensembles of Local Experts in Metabonomic Applications

MCS '09 Proceedings of the 8th International Workshop on Multiple Classifier Systems
Tree Covering within a Graph Kernel Framework for Shape Classification

ICIAP '09 Proceedings of the 15th International Conference on Image Analysis and Processing
Reducing the run-time complexity of support vector data descriptions

IJCNN'09 Proceedings of the 2009 international joint conference on Neural Networks
Combining SOM and local minimum enclosing spheres for novelty detection

CCDC'09 Proceedings of the 21st annual international conference on Chinese control and decision conference
Fast support vector data descriptions for novelty detection

IEEE Transactions on Neural Networks
A Computable Plug-In Estimator of Minimum Volume Sets for Novelty Detection

Operations Research
Feature extraction for novelty detection as applied to fault detection in machinery

Pattern Recognition Letters
Front end analysis of speech recognition: a review

International Journal of Speech Technology
People re-identification by graph kernels methods

GbRPR'11 Proceedings of the 8th international conference on Graph-based representations in pattern recognition
A graph-kernel method for re-identification

ICIAR'11 Proceedings of the 8th international conference on Image analysis and recognition - Volume Part I
Variable window adaptive Kernel Principal Component Analysis for nonlinear nonstationary process monitoring

Computers and Industrial Engineering
Kernel PCA in application to leakage detection in drinking water distribution system

ICCCI'11 Proceedings of the Third international conference on Computational collective intelligence: technologies and applications - Volume Part I
PCA document reconstruction for email classification

Computational Statistics & Data Analysis
Novelty detection for the inspection of light-emitting diodes

Expert Systems with Applications: An International Journal
A Closed-form Solution for the Pre-image Problem in Kernel-based Machines

Journal of Signal Processing Systems
Geometrically local embedding in manifolds for dimension reduction

Pattern Recognition
Meta-cognitive Neural Network for classification problems in a sequential learning framework

Neurocomputing
On feature selection with principal component analysis for one-class SVM

Pattern Recognition Letters
Feature selection from high-order tensorial data via sparse decomposition

Pattern Recognition Letters
A meta-cognitive sequential learning algorithm for neuro-fuzzy inference system

Applied Soft Computing
L1 norm based KPCA for novelty detection

Pattern Recognition
Meta-cognitive RBF Network and its Projection Based Learning algorithm for classification problems

Applied Soft Computing
A modified support vector data description based novelty detection approach for machinery components

Applied Soft Computing
Novelty detection using a new group outlier factor

ICONIP'12 Proceedings of the 19th international conference on Neural Information Processing - Volume Part III
Document categorization based on minimum loss of reconstruction information

MICAI'12 Proceedings of the 11th Mexican international conference on Advances in Computational Intelligence - Volume Part II
One-class classification with Gaussian processes

Pattern Recognition
Minimizer of the Reconstruction Error for multi-class document categorization

Expert Systems with Applications: An International Journal
Geometrical and computational aspects of Spectral Support Estimation for novelty detection

Pattern Recognition Letters
Review: A review of novelty detection

Signal Processing

Quantified Score

Hi-index	0.01

Visualization

Abstract

Kernel principal component analysis (kernel PCA) is a non-linear extension of PCA. This study introduces and investigates the use of kernel PCA for novelty detection. Training data are mapped into an infinite-dimensional feature space. In this space, kernel PCA extracts the principal components of the data distribution. The squared distance to the corresponding principal subspace is the measure for novelty. This new method demonstrated a competitive performance on two-dimensional synthetic distributions and on two real-world data sets: handwritten digits and breast-cancer cytology.