Generalized Correntropy for Robust Adaptive Filtering

Abstract

As a robust nonlinear similarity measure in kernel space, correntropy has received increasing attention in domains of machine learning and signal processing. In particular, the maximum correntropy criterion (MCC) has recently been successfully applied in robust regression and filtering. The default kernel function in correntropy is the Gaussian kernel, which is, of course, not always the best choice. In this work, we propose a generalized correntropy that adopts the generalized Gaussian density (GGD) function as the kernel (not necessarily a Mercer kernel), and present some important properties. We further propose the generalized maximum correntropy criterion (GMCC), and apply it to adaptive filtering. An adaptive algorithm, called the GMCC algorithm, is derived, and the mean square convergence performance is studied. We show that the proposed algorithm is very stable and can achieve zero probability of divergence (POD). Simulation results confirm the theoretical expectations and demonstrate the desirable performance of the new algorithm.

34 pages, 9 figures, submitted to IEEE Transactions on Signal Processing

Cited by in corpus (24)

• Kernel Risk-Sensitive Loss: Definition, Properties and Application to Robust Adaptive Filtering
• Maximum Correntropy Criterion with Variable Center
• Robust Matrix Completion via Maximum Correntropy Criterion and Half Quadratic Optimization
• Generalized Multi-kernel Maximum Correntropy Kalman Filter for Disturbance Estimation
• Study of Robust Sparsity-Aware RLS algorithms with Jointly-Optimized Parameters for Impulsive Noise Environments
• Learning with Correntropy-induced Losses for Regression with Mixture of Symmetric Stable Noise
• A Statistical Learning Approach to Modal Regression
• Robust Tensor Decomposition for Image Representation Based on Generalized Correntropy
• Online Nonlinear Estimation via Iterative L2-Space Projections: Reproducing Kernel of Subspace
• RELF: Robust Regression Extended with Ensemble Loss Function
• Asymmetric Correntropy for Robust Adaptive Filtering
• Adaptive sparseness for correntropy-based robust regression via automatic relevance determination
• Robust Low-tubal-rank Tensor Completion based on Tensor Factorization and Maximum Correntopy Criterion
• Fractional-Order Subband p-Norm Adaptive Filter via Transformation Nearest Kronecker Product Decomposition for Active Noise Control
• Proportionate Adaptive Filtering under Correntropy Criterion in Impulsive Noise Environments
• Sparsity-Aware SSAF Algorithm with Individual Weighting Factors for Acoustic Echo Cancellation
• A Robust Matching Pursuit Algorithm Using Information Theoretic Learning
• A Unified Weight Learning and Low-Rank Regression Model for Robust Complex Error Modeling
• Study of Diffusion Normalized Least Mean M-estimate Algorithms
• Study of Proximal Normalized Subband Adaptive Algorithm for Acoustic Echo Cancellation
• Robust Motion Averaging under Maximum Correntropy Criterion
• Mean-Square Performance Analysis of Noise-Robust Normalized Subband Adaptive Filter Algorithm
• Marine Animal Classification with Correntropy Loss Based Multi-view Learning
• New Insights into Learning with Correntropy Based Regression