Learning with correntropy-induced losses for regression with mixture of symmetric stable noise
From MaRDI portal
Publication:2300760
Abstract: In recent years, correntropy and its applications in machine learning have been drawing continuous attention owing to its merits in dealing with non-Gaussian noise and outliers. However, theoretical understanding of correntropy, especially in the statistical learning context, is still limited. In this study, within the statistical learning framework, we investigate correntropy based regression in the presence of non-Gaussian noise or outliers. Motivated by the practical way of generating non-Gaussian noise or outliers, we introduce mixture of symmetric stable noise, which include Gaussian noise, Cauchy noise, and their mixture as special cases, to model non-Gaussian noise or outliers. We demonstrate that under the mixture of symmetric stable noise assumption, correntropy based regression can learn the conditional mean function or the conditional median function well without resorting to the finite-variance or even the finite first-order moment condition on the noise. In particular, for the above two cases, we establish asymptotic optimal learning rates for correntropy based regression estimators that are asymptotically of type . These results justify the effectiveness of the correntropy based regression estimators in dealing with outliers as well as non-Gaussian noise. We believe that the present study completes our understanding towards correntropy based regression from a statistical learning viewpoint, and may also shed some light on robust statistical learning for regression.
Recommendations
- Learning with the maximum correntropy criterion induced losses for regression
- New insights into learning with correntropy-based regression
- Optimal learning with Gaussians and correntropy loss
- Smooth -insensitive regression by loss symmetrization
- Smooth -insensitive regression by loss symmetrization.
- scientific article; zbMATH DE number 6253925
- A robust classification framework with mixture correntropy
- Settling the robust learnability of mixtures of Gaussians
- Kernel-based sparse regression with the correntropy-induced loss
Cites work
- A Statistical Learning Approach to Modal Regression
- Concentration estimates for learning with \(\ell ^{1}\)-regularizer and data dependent hypothesis spaces
- Concentration estimates for learning with unbounded sampling
- Consistency analysis of an empirical minimum error entropy algorithm
- Correntropy: Properties and Applications in Non-Gaussian Signal Processing
- Data mining. Concepts and techniques
- Finite-memory denoising in impulsive noise using Gaussian mixture models
- Generalized correlation function: definition, properties, and application to blind equalization
- Generalized Correntropy for Robust<?Pub _newline ?>Adaptive Filtering
- scientific article; zbMATH DE number 3954047 (Why is no real title available?)
- scientific article; zbMATH DE number 3684698 (Why is no real title available?)
- scientific article; zbMATH DE number 1332320 (Why is no real title available?)
- scientific article; zbMATH DE number 614990 (Why is no real title available?)
- scientific article; zbMATH DE number 621792 (Why is no real title available?)
- scientific article; zbMATH DE number 194744 (Why is no real title available?)
- scientific article; zbMATH DE number 194758 (Why is no real title available?)
- scientific article; zbMATH DE number 5018128 (Why is no real title available?)
- scientific article; zbMATH DE number 3320125 (Why is no real title available?)
- Information theoretic learning. Renyi's entropy and kernel perspectives
- Learning rates of least-square regularized regression
- Learning Theory
- Learning theory approach to minimum error entropy criterion
- Learning with the maximum correntropy criterion induced losses for regression
- Learning without concentration
- Maximum correntropy Kalman filter
- Multi-kernel regularized classifiers
- On the behavior of Tukey's depth and median under symmetric stable distributions.
- On-line Bayesian estimation of signals in symmetric /spl alpha/-stable noise
- Optimal learning rates for least squares regularized regression with unbounded sampling
- Outlier analysis.
- Parameter Estimates for Symmetric Stable Distributions
- Properties of certain symmetric stable distributions
- Robust Estimation of a Location Parameter
- Robust Hyperspectral Unmixing With Correntropy-Based Metric
- Robust Principal Component Analysis Based on Maximum Correntropy Criterion
- Robust Statistics
- Robust techniques for signal processing: A survey
- Soft-Decision Metrics for Coded Orthogonal Signaling in Symmetric Alpha-Stable Noise
- Stable Distributions in Statistical Inference: 1. Symmetric Stable Distributions Compared to Other Symmetric Long-Tailed Distributions
- Statistical analysis of finite mixture distributions
- Support Vector Machines
- The C-loss function for pattern classification
- The stability test for symmetric alpha-stable distributions
Cited in
(20)- Fast rates of minimum error entropy with heavy-tailed noise
- Kernel-based maximum correntropy criterion with gradient descent method
- A robust classification framework with mixture correntropy
- Kernel-based sparse regression with the correntropy-induced loss
- Error analysis on regularized regression based on the maximum correntropy criterion
- A Statistical Learning Approach to Modal Regression
- New insights into learning with correntropy-based regression
- A Framework of Learning Through Empirical Gain Maximization
- Robust wavelet-based estimation for varying coefficient dynamic models under long-dependent structures
- Analysis of the SNR Loss Distribution With Covariance Mismatched Training Samples
- A regularized correntropy framework for robust pattern recognition
- Correntropy: Implications of nonGaussianity for the moment expansion and deconvolution
- Learning with the maximum correntropy criterion induced losses for regression
- Optimal learning with Gaussians and correntropy loss
- Maximum correntropy criterion regression models with tending-to-zero scale parameters
- Maximum correntropy regression model with mixture of triangular noise
- Fast rates of exponential cost function
- Tikhonov regularization for Gaussian empirical gain maximization in RKHS is consistent
- Learning theory of regularized Huber regression
- Learning under \((1 + \epsilon)\)-moment conditions
This page was built for publication: Learning with correntropy-induced losses for regression with mixture of symmetric stable noise
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2300760)
