arXiv subjects

Gerlind Plonka

Publications and source records attributed to Gerlind Plonka.

27 records · Page 2Linked to original sources

arXiv2020

Modifications of Prony's Method for the Recovery and Sparse Approximation of Generalized Exponential Sums

In this survey we describe some modifications of Prony's method. In particular, we consider the recovery of general expansions into eigenfunctions of linear differential operators of first order and show, how these expansions can be recovered from function samples using generalized shift operators. We derive an ESPRIT-like algorithm for the generalized recovery method and show, that this approach can be directly used to reconstruct classical exponential sums from non-equispaced data. Furthermore, we derive a modification of Prony's method for sparse approximation with exponential sums which leads to a non-linear least-squares problem.

math.NA↗

arXiv2019

Frame Soft Shrinkage as Proximity Operator

Let $\mathcal H$ and $\mathcal K$ be real Hilbert spaces and $T \in \mathcal{B} (\mathcal H,\mathcal K)$ an injective operator with closed range and Moore-Penrose inverse $T^\dagger$. Based on the well-known characterization of proximity operators by Moreau, we prove that for any proximity operator $\text{Prox} \colon \mathcal K \to \mathcal K$ the operator $T^\dagger \, \text{Prox} \, T$ is a proximity operator on the linear space $\mathcal H$ equipped with a suitable norm. In particular, it follows for the frequently applied soft shrinkage operator $\text{Prox} = S_λ\colon \ell_2 \rightarrow \ell_2$ and any frame analysis operator $T\colon \mathcal H \to \ell_2$, that the frame shrinkage operator $T^\dagger\, S_λ\, T$ is a proximity operator in a suitable Hilbert space.

math.OC↗

arXiv2018

Real Sparse Fast DCT for Vectors with Short Support

In this paper we present a new fast and deterministic algorithm for the inverse discrete cosine transform of type II for reconstructing the input vector $\mathbf x\in\mathbb R^N$, $N=2^J$, with short support of length $m$ from its discrete cosine transform $\mathbf x^{\widehat{\mathrm{II}}}=C^{\mathrm{II}}_N\mathbf x$ if an upper bound $M\geq m$ is known. The resulting algorithm only uses real arithmetic, has a runtime of $\mathcal{O}\left(M\log M+m\log_2\frac{N}{M}\right)$ and requires $\mathcal{O}\left(M+m\log_2\frac{N}{M}\right)$ samples of $\mathbf x^{\widehat{\mathrm{II}}}$. For $m,M\rightarrow N$ the runtime and sampling requirements approach those of a regular IDCT-II for vectors with full support. The algorithm presented hereafter does not employ inverse FFT algorithms to recover $\mathbf x$.

math.NA↗

arXiv2018

Sparse Fast DCT for Vectors with One-block Support

In this paper we present a new fast and deterministic algorithm for the inverse discrete cosine transform of type II that reconstructs the input vector $\mathbf{x}\in\mathbb{R}^{N}$, $N=2^{J-1}$, with short support of length $m$ from its discrete cosine transform $\mathbf{x}^{\widehat{\mathrm{II}}}=\mathbf{C}_N^{\mathrm{II}}\mathbf{x}$. The resulting algorithm has a runtime of $\mathcal{O}\left(m\log m\log \frac{2N}{m}\right)$ and requires $\mathcal{O}\left(m\log \frac{2N}{m}\right)$ samples of $\mathbf{x}^{\widehat{\mathrm{II}}}$. In order to derive this algorithm we also develop a new fast and deterministic inverse FFT algorithm that constructs the input vector $\mathbf{y}\in\mathbb{R}^{2N}$ with reflected block support of block length $m$ from $\widehat{\mathbf{y}}$ with the same runtime and sampling complexities as our DCT algorithm.

math.NA↗

arXiv2017

Sparse phase retrieval of one-dimensional signals by Prony's method

In this paper, we show that sparse signals f representable as a linear combination of a finite number N of spikes at arbitrary real locations or as a finite linear combination of B-splines of order m with arbitrary real knots can be almost surely recovered from O(N^2) Fourier intensity measurements up to trivial ambiguities. The constructive proof consists of two steps, where in the first step the Prony method is applied to recover all parameters of the autocorrelation function and in the second step the parameters of f are derived. Moreover, we present an algorithm to evaluate f from its Fourier intensities and illustrate it at different numerical examples.

math.NA↗

arXiv2016

Application of the AAK theory for sparse approximation of exponential sums

In this paper, we derive a new method for optimal $\ell^{1}$- and $\ell^2$-approximation of discrete signals on ${\mathbb N}_{0}$ whose entries can be represented as an exponential sum of finite length. Our approach employs Prony's method in a first step to recover the exponential sum that is determined by the signal. In the second step we use the AAK-theory to derive an algorithm for computing a shorter exponential sum that approximates the original signal in the $\ell^{p}$-norm well. AAK-theory originally determines best approximations of bounded periodic functions in Hardy-subspaces. We rewrite these ideas for our purposes and give a proof of the used AAK theorem based only on basic tools from linear algebra and Fourier analysis. The new algorithm is tested numerically in different examples.

math.NA↗

arXiv2016

Enforcing uniqueness in one-dimensional phase retrieval by additional signal information in time domain

Considering the ambiguousness of the discrete-time phase retrieval problem to recover a signal from its Fourier intensities, one can ask the question: what additional information about the unknown signal do we need to select the correct solution within the large solution set? Based on a characterization of the occurring ambiguities, we investigate different a priori conditions in order to reduce the number of ambiguities or even to receive a unique solution. Particularly, if we have access to additional magnitudes of the unknown signal in the time domain, we can show that almost all signals with finite support can be uniquely recovered. Moreover, we prove that an analogous result can be obtained by exploiting additional phase information.

math.NA↗

arXiv2015

Pseudo-inverses of difference matrices and their application to sparse signal approximation

We derive new explicit expressions for the components of Moore-Penrose inverses of symmetric difference matrices. These generalized inverses are applied in a new regularization approach for scattered data interpolation based on partial differential equations. The columns of the Moore-Penrose inverse then serve as elements of a dictionary that allow a sparse signal approximation. In order to find a set of suitable data points for signal representation we apply the orthogonal patching pursuit (OMP) method.

math.NA↗

arXiv2015

A deterministic sparse FFT algorithm for vectors with small support

In this paper we consider the special case where a discrete signal ${\bf x}$ of length N is known to vanish outside a support interval of length $m < N$. If the support length $m$ of ${\bf x}$ or a good bound of it is a-priori known we derive a sublinear deterministic algorithm to compute ${\bf x}$ from its discrete Fourier transform. In case of exact Fourier measurements we require only ${\cal O}(m \log m)$ arithmetical operations. For noisy measurements, we propose a stable ${\cal O}(m \log N)$ algorithm.

math.NA↗