Generalized mean for robust principal component analysis

In this paper, we propose a robust principal component analysis (PCA) to overcome the problem that PCA is prone to outliers included in the training set. Different from the other alternatives which commonly replace L2-norm by other distance measures, the proposed method alleviates the negative effect of outliers using the characteristic of the generalized mean keeping the use of the Euclidean distance. The optimization problem based on the generalized mean is solved by a novel method. We also present a generalized sample mean, which is a generalization of the sample mean, to estimate a robust mean in the presence of outliers. The proposed method shows better or equivalent performance than the conventional PCAs in various problems such as face reconstruction, clustering, and object categorization. HighlightsWe propose a robust principal component analysis.The generalized mean is used in the proposed method instead of the arithmetic mean.A novel method is also presented to solve our optimization problem.

• Publication year

  2016

• Venue

  Pattern Recognition

• Publication date

  2016-06-01

• Fields of study

  Mathematics, Computer Science

• Identifiers
  DOI 10.1016/j.patcog.2016.01.002
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Feature Selection

  2019cited by this paper
• Robust Statistics

  2018cited by this paper
• Grassmann Averages for Scalable Robust PCA

  2014cited by this paper
• Principal Component Analysis by $L_{p}$ -Norm Maximization

  2014cited by this paper
• Generalized mean for feature extraction in one-class classification problems

  2013cited by this paper
• Feature extraction based on Lp-norm generalized principal component analysis

  2013cited by this paper
• A pure L1L1-norm principal component analysis

  2013cited by this paper
• A Pure L1-norm Principal Component Analysis.

  2013cited by this paper
• Robust Principal Component Analysis Based on Maximum Correntropy Criterion

  2011influential reference
• Principal component analysis based on non-parametric maximum entropy

  2010cited by this paper
• Robust principal component analysis?

  2009cited by this paper
• Principal Component Analysis Based on L1-Norm Maximization

  2008influential reference
• Incremental Learning for Robust Visual Tracking

  2008cited by this paper
• Correntropy: Properties and Applications in Non-Gaussian Signal Processing

  2007cited by this paper
• R1-PCA: rotational invariant L1-norm principal component analysis for robust subspace factorization

  2006influential reference
• Pattern Recognition and Machine Learning

  2006cited by this paper
• Pattern Recognition

  2005cited by this paper
• Robust L/sub 1/ norm factorization in the presence of outliers and missing data by alternative convex programming

  2005influential reference
• Elements of Information Theory

  2005cited by this paper
• Feature selection, L1 vs. L2 regularization, and rotational invariance

  2004cited by this paper
• Learning methods for generic object recognition with invariance to pose and lighting

  2004cited by this paper
• K-means clustering via principal component analysis

  2004cited by this paper
• Principal Component Analysis

  2003cited by this paper
• A Framework for Robust Subspace Learning

  2003cited by this paper
• Mean Shift: A Robust Approach Toward Feature Space Analysis

  2002cited by this paper
• Principal component analysis for clustering gene expression data

  2001cited by this paper
• Statistical Pattern Recognition: A Review

  2000cited by this paper
• The FERET evaluation methodology for face-recognition algorithms

  1997cited by this paper
• Eigenfaces vs. Fisherfaces: Recognition Using Class Specific Linear Projection

  1996cited by this paper
• Eigenfaces for Recognition

  1991influential reference
• Handbook of means and their inequalities

  1987cited by this paper
• Matrix computations

  1983cited by this paper

• Outlier-Robust Learning with Continuously Differentiable Least Trimmed Squares

  2026cites this paper
• Two new multi-view robust principal component analysis methods

  2026cites this paper
• Data Subdivision Based Dual-Weighted Robust Principal Component Analysis

  2025cites this paper
• Reweighted angle two-dimensional principal component analysis for feature extraction

  2025cites this paper
• Maximizing generalized mean of trace ratios for discriminative feature learning

  2025cites this paper
• Cancelable biometric authentication leveraging empirical mode decomposition and quaternion representations for IoT security

  2025cites this paper
• Enhanced intrusion detection through dual reduction and robust mean

  2025cites this paper
• Using Maximum Weighted Likelihood to Derive Lehmer and Hölder Families of Means

  2024cites this paper
• Deriving Lehmer and Hölder means as maximum weighted likelihood estimates for the multivariate exponential family

  2024cites this paper
• Robust multi-view locality preserving regression embedding

  2024cites this paper
• On permutation-invariant neural networks

  2024cites this paper
• Efficient Dimensionality Reduction in Intrusion Detection Systems via Weighted Power Mean for PCA and LDA

  2024cites this paper
• Using maximum weighted likelihood to derive Lehmer and H\"older mean families

  2023cites this paper
• Agreeing to Stop: Reliable Latency-Adaptive Decision Making via Ensembles of Spiking Neural Networks

  2023cites this paper
• A robust intrusion detection model based on a combination of PCA-GM and trunced LDA

  2023cites this paper
• Avoiding Optimal Mean Robust and Sparse BPCA with L1-norm Maximization

  2023cites this paper
• Discrete Robust Principal Component Analysis via Binary Weights Self-Learning

  2022cites this paper
• Power mean based image segmentation in the presence of noise

  2022influential citation
• Unified feature extraction framework based on contrastive learning

  2021influential citation
• Uncertainty-aware Generalized Adaptive CycleGAN

  2021cites this paper
• Self-paced Principal Component Analysis

  2021cites this paper
• Robustness via Uncertainty-aware Cycle Consistency

  2021cites this paper
• A Unified Framework for Feature Extraction based on Contrastive Learning

  2021cites this paper
• A New Active Contours Image Segmentation Model Driven by Generalized Mean with Outlier Restoration Achievements

  2020cites this paper
• Joint low-rank project embedding and optimal mean principal component analysis

  2020cites this paper
• Robust deflated principal component analysis via multiple instance factorings for dimension reduction in remote sensing images

  2020cites this paper
• Deflated manifold embedding PCA framework via multiple instance factorings

  2020cites this paper
• A NEW NON-CONVEX APPROACH FOR COMPRESSIVE SENSING MRI

  2020cites this paper
• Robust Sparse Representation and Multiclass Support Matrix Machines for the Classification of Motor Imagery EEG Signals

  2019cites this paper
• A generalized mean distance-based k-nearest neighbor classifier

  2019cites this paper
• Efficient robust methods via monitoring for clustering and multivariate data analysis

  2019cites this paper
• Robust High-Order Manifold Constrained Sparse Principal Component Analysis for Image Representation

  2019influential citation
• Optimization of blasting parameters for an underground mine through prediction of blasting vibration

  2019cites this paper
• A generalized multi-dictionary least squares framework regularized with multi-graph embeddings

  2019cites this paper
• Improve R2L Attack Detection Using Trimmed PCA

  2019cites this paper
• Two-dimensional locality adaptive discriminant analysis

  2019cites this paper
• A cyber network attack detection based on GM Median Nearest Neighbors LDA

  2019cites this paper
• Sparse Representation and Support Matrix Machines for Epileptic EEG Signal Classification

  2019cites this paper
• Robust Sparse Representation and Multiclass Support Matrix Machines for the Classification of Motor Imagery EEG Signals

  2019cites this paper
• Robust Formulation for PCA: Avoiding Mean Calculation With L2, p-norm Maximization

  2018cites this paper
• Invariance properties for the error function used for multilinear regression

  2018cites this paper
• Robust PCAs and PCA Using Generalized Mean

  2018cites this paper
• An accurate traffic classification model based on support vector machines

  2017cites this paper
• Linguistically Described Covariance Matrix Estimation

  2017cites this paper
• Two-Dimensional PCA with F-Norm Minimization

  2017cites this paper
• Regularized coplanar discriminant analysis for dimensionality reduction

  2017cites this paper
• Avoiding Optimal Mean ℓ2,1-Norm Maximization-Based Robust PCA for Reconstruction

  2017cites this paper
• Optimal mean two-dimensional principal component analysis with F-norm minimization

  2017cites this paper
• Identification of pesticide residue level in lettuce based on hyperspectra and chlorophyll fluorescence spectra

  2016cites this paper
• Building and generating facial textures using Eigen faces

  2016cites this paper
• Robust Learning With Kernel Mean $p$ -Power Error Loss

  2016influential citation