Geometrical Study of Matching Pursuit Parametrization
ثبت نشده
چکیده
This paper studies the effect of discretizing the parametrization of a dictionary used for Matching Pursuit decompositions of signals. Our approach relies on viewing the continuously parametrized dictionary as an embedded manifold in the signal space on which the tools of differential (Riemannian) geometry can be applied. The main contribution of this paper is twofold. First, we prove that if a discrete dictionary reaches a minimal density criterion, then the corresponding discrete MP (dMP) is equivalent in terms of convergence to a weakened hypothetical continuous MP. Interestingly, the corresponding weakness factor depends on a density measure of the discrete dictionary. Second, we show that the insertion of a simple geometric gradient ascent optimization on the atom dMP selection maintains the previous comparison but with a weakness factor at least two times closer to unity than without optimization. Finally, we present numerical experiments confirming our theoretical predictions for decomposition of signals and images on regular discretizations of dictionary parametrizations.
منابع مشابه
A Unified Time-Frequency Parametrization of EEGs The Matching Pursuit Algorithm Encompasses Properties of Spectral and Visual EEG Analysis
Seventy years since the first recording of the human electroencephalogram (EEG) [1], visual analysis of raw EEG traces is still the major clinical tool and point of reference for other methods, in spite of its inherent limitations: low repeatability and high cost. Seven years since the introduction of the matching pursuit (MP) [2] algorithm, we have collected evidence suggesting that adaptive t...
متن کاملFractional Spline Matching Pursuit: a quantitative tool for seismic stratigraphy
A matching pursuit algorithm is used to obtain quantitative information on the litho-stratigraphical content of seismic data. The method is based on a greedy non-linear algorithm decomposing the data into atoms drawn from a redundant dictionary defined by fractional splines. The atom’s parametrization provides localized scale, order and direction attributes which carry information on the strati...
متن کاملPMU-Based Matching Pursuit Method for Black-Box Modeling of Synchronous Generator
This paper presents the application of the matching pursuit method to model synchronous generator. This method is useful for online analysis. In the proposed method, the field voltage is considered as input signal, while the terminal voltage and active power of the generator are output signals. Usually, the difference equation with a second degree polynomial structure is used to estimate the co...
متن کاملFrom wavelets to adaptive approximations: time-frequency parametrization of EEG
This paper presents a summary of time-frequency analysis of the electrical activity of the brain (EEG). It covers in details two major steps: introduction of wavelets and adaptive approximations. Presented studies include time-frequency solutions to several standard research and clinical problems, encountered in analysis of evoked potentials, sleep EEG, epileptic activities, ERD/ERS and pharmac...
متن کامل