-
Book
-
Chapter
Introduction
Robust data classification or representation is a fundamental task and has a long history in computer vision. The algorithmic robustness, which is derived from the statistical definition of a breakdown point [...
-
Chapter
M-Estimators and Half-Quadratic Minimization
In robust statistics, there are several types of robust estimators, including M-estimator (maximum likelihood type estimator), L-estimator (linear combinations of order statistics), R-estimator (estimator base...
-
Chapter
Correntropy and Linear Representation
The nearest neighbor (NN) classifier is the most popular method for image-based object recognition. In NN classifier, the representational capacity of an image database and the recognition rate depend on how r...
-
Chapter
Correntropy with Nonnegative Constraint
Nonnegativity constraint is more consistent with the biological modeling of visual data and often leads to better performance for data representation and graph learning [66]. In this chapter, we present an ove...
-
Chapter
Information Measures
Information theoretic learning (ITL) was initiated in the late 1990s at CNEL [126]. It uses descriptors from information theory (entropy and divergences) estimated directly from the data to substitute the conv...
-
Chapter
ℓ 1 Regularized Correntropy
Sparse signal representation arises in application of compressed sensing and has been considered as a significant technique in computer vision and machine learning [27, 65, 154]. Based on the ℓ 0-ℓ ...