Searcharxiv⌕ Search

arXiv subjects

Rachel Ward

Publications and source records attributed to Rachel Ward.

At least 91 records · Page 5Linked to original sources
arXiv2012

A comparison of the discrete Kolmogorov-Smirnov statistic and the Euclidean distance

Goodness-of-fit tests gauge whether a given set of observations is consistent (up to expected random fluctuations) with arising as independent and identically distributed (i.i.d.) draws from a user-specified probability distribution known as the "model." The standard gauges involve the discrepancy between the model and the empirical distribution of the observed draws. Some measures of discrepancy are cumulative; others are not. The most popular cumulative measure is the Kolmogorov-Smirnov statistic; when all probability distributions under consideration are discrete, a natural noncumulative measure is the Euclidean distance between the model and the empirical distributions. In the present paper, both mathematical analysis and its illustration via various data sets indicate that the Kolmogorov-Smirnov statistic tends to be more powerful than the Euclidean distance when there is a natural ordering for the values that the draws can take -- that is, when the data is ordinal -- whereas the Euclidean distance is more reliable and more easily understood than the Kolmogorov-Smirnov statistic when there is no natural ordering (or partial order) -- that is, when the data is nominal.

stat.ME↗
arXiv2012

Stability for second-order chaotic sigma delta quantization

We prove that that second-order (double-loop) chaotic sigma-delta schemes are stable - within a certain parameter range, all state variables of the system are guaranteed to remain uniformly bounded. To our knowledge this is the first general stability result for chaotic sigma-delta schemes of order greater than one. Invariably as the amount of expansion added to the system is increased, the dynamic range of the input must get smaller for stability to be guaranteed. We give explicit bounds on this trade-off and verify through numerical simulation that these bounds are near-optimal.

math.DS↗
arXiv2012

Two-subspace Projection Method for Coherent Overdetermined Systems (Technical Report)

In this technical report we present a Projection onto Convex Sets (POCS) type algorithm for solving systems of linear equations. POCS methods have found many applications ranging from computer tomography to digital signal and image processing. The Kaczmarz method is one of the most popular solvers for overdetermined systems of linear equations due to its speed and simplicity. Here we introduce and analyze an extension of the Kaczmarz method which iteratively projects the estimate onto a solution space given from two randomly selected rows. We show that this projection algorithm provides exponential convergence to the solution in expectation. The convergence rate significantly improves upon that of the standard randomized Kaczmarz method when the system has coherent rows. We also show that the method is robust to noise, and converges exponentially in expectation to the noise floor. Experimental results are provided which confirm that in the coherent case our method significantly outperforms the randomized Kaczmarz method.

math.NA↗
arXiv2012

An introduction to how chi-square and classical exact tests often wildly misreport significance and how the remedy lies in computers

Goodness-of-fit tests based on the Euclidean distance often outperform chi-square and other classical tests (including the standard exact tests) by at least an order of magnitude when the model being tested for goodness-of-fit is a discrete probability distribution that is not close to uniform. The present article discusses numerous examples of this. Goodness-of-fit tests based on the Euclidean metric are now practical and convenient: although the actual values taken by the Euclidean distance and similar goodness-of-fit statistics are seldom humanly interpretable, black-box computer programs can rapidly calculate their precise significance.

stat.ME↗
arXiv2011

Computing the confidence levels for a root-mean-square test of goodness-of-fit, II

This paper extends our earlier article, "Computing the confidence levels for a root-mean-square test of goodness-of-fit;" unlike in the earlier article, the models in the present paper involve parameter estimation -- both the null and alternative hypotheses in the associated tests are composite. We provide efficient black-box algorithms for calculating the asymptotic confidence levels of a variant on the classic chi-squared test. In some circumstances, it is also feasible to compute the exact confidence levels via Monte Carlo simulation.

stat.CO↗
arXiv2011

Weighted eigenfunction estimates with applications to compressed sensing

Using tools from semiclassical analysis, we give weighted L^\infty estimates for eigenfunctions of strictly convex surfaces of revolution. These estimates give rise to new sampling techniques and provide improved bounds on the number of samples necessary for recovering sparse eigenfunction expansions on surfaces of revolution. On the sphere, our estimates imply that any function having an s-sparse expansion in the first N spherical harmonics can be efficiently recovered from its values at m > s N^(1/6) log^4(N) sampling points.

math.NA↗
arXiv2011

Chi-square and classical exact tests often wildly misreport significance; the remedy lies in computers

If a discrete probability distribution in a model being tested for goodness-of-fit is not close to uniform, then forming the Pearson chi-square statistic can involve division by nearly zero. This often leads to serious trouble in practice -- even in the absence of round-off errors -- as the present article illustrates via numerous examples. Fortunately, with the now widespread availability of computers, avoiding all the trouble is simple and easy: without the problematic division by nearly zero, the actual values taken by goodness-of-fit statistics are not humanly interpretable, but black-box computer programs can rapidly calculate their precise significance.

stat.ME↗
arXiv2011

Low-rank matrix recovery via iteratively reweighted least squares minimization

We present and analyze an efficient implementation of an iteratively reweighted least squares algorithm for recovering a matrix from a small number of linear measurements. The algorithm is designed for the simultaneous promotion of both a minimal nuclear norm and an approximatively low-rank solution. Under the assumption that the linear measurements fulfill a suitable generalization of the Null Space Property known in the context of compressed sensing, the algorithm is guaranteed to recover iteratively any matrix with an error of the order of the best k-rank approximation. In certain relevant cases, for instance for the matrix completion problem, our version of this algorithm can take advantage of the Woodbury matrix identity, which allows to expedite the solution of the least squares problems required at each iteration. We present numerical experiments that confirm the robustness of the algorithm for the solution of matrix completion problems, and demonstrate its competitiveness with respect to other techniques proposed recently in the literature.

math.NA↗
arXiv2011

Sparse Legendre expansions via $\ell_1$ minimization

We consider the problem of recovering polynomials that are sparse with respect to the basis of Legendre polynomials from a small number of random samples. In particular, we show that a Legendre s-sparse polynomial of maximal degree N can be recovered from m = O(s log^4(N)) random samples that are chosen independently according to the Chebyshev probability measure. As an efficient recovery method, l1-minimization can be used. We establish these results by verifying the restricted isometry property of a preconditioned random Legendre matrix. We then extend these results to a large class of orthogonal polynomial systems, including the Jacobi polynomials, of which the Legendre polynomials are a special case. Finally, we transpose these results into the setting of approximate recovery for functions in certain infinite-dimensional function spaces.

math.NA↗
arXiv2011

Computing the confidence levels for a root-mean-square test of goodness-of-fit

The classic chi-squared statistic for testing goodness-of-fit has long been a cornerstone of modern statistical practice. The statistic consists of a sum in which each summand involves division by the probability associated with the corresponding bin in the distribution being tested for goodness-of-fit. Typically this division should precipitate rebinning to uniformize the probabilities associated with the bins, in order to make the test reasonably powerful. With the now widespread availability of computers, there is no longer any need for this. The present paper provides efficient black-box algorithms for calculating the asymptotic confidence levels of a variant on the classic chi-squared test which omits the problematic division. In many circumstances, it is also feasible to compute the exact confidence levels via Monte Carlo simulation.

stat.CO↗
arXiv2011

Sparse recovery for spherical harmonic expansions

We show that sparse spherical harmonic expansions can be efficiently recovered from a small number of randomly chosen samples on the sphere. To establish the main result, we verify the restricted isometry property of an associated preconditioned random measurement matrix using recent estimates on the uniform growth of Jacobi polynomials.

math.NA↗
arXiv2011

New and improved Johnson-Lindenstrauss embeddings via the Restricted Isometry Property

Consider an m by N matrix Phi with the Restricted Isometry Property of order k and level delta, that is, the norm of any k-sparse vector in R^N is preserved to within a multiplicative factor of 1 +- delta under application of Phi. We show that by randomizing the column signs of such a matrix Phi, the resulting map with high probability embeds any fixed set of p = O(e^k) points in R^N into R^m without distorting the norm of any point in the set by more than a factor of 1 +- delta. Consequently, matrices with the Restricted Isometry Property and with randomized column signs provide optimal Johnson-Lindenstrauss embeddings up to logarithmic factors in N. In particular, our results improve the best known bounds on the necessary embedding dimension m for a wide class of structured random matrices; for partial Fourier and partial Hadamard matrices, we improve the recent bound m = O(delta^(-4) log(p) log^4(N)) appearing in Ailon and Liberty to m = O(delta^(-2) log(p) log^4(N)), which is optimal up to the logarithmic factors in N. Our results also have a direct application in the area of compressed sensing for redundant dictionaries.

cs.IT↗
arXiv2010

Quiet sigma delta quantization, and global convergence for a class of asymmetric piecewise affine maps

In this paper, we introduce a family of second-order sigma delta quantization schemes for analog-to-digital conversion which are `quiet' : quantization output is guaranteed to fall to zero at the onset of vanishing input. In the process, we prove that the origin is a globally attractive fixed point for the related family of asymmetrically-damped piecewise affine maps. Our proof of convergence is twofold: first, we construct a trapping set using a Lyapunov-type argument; we then take advantage of the asymmetric structure of the maps under consideration to prove convergence to the origin from within this trapping set.

math.DS↗
arXiv2010

Lower bounds for the error decay incurred by coarse quantization schemes

Several analog-to-digital conversion methods for bandlimited signals used in applications, such as Sigma Delta quantization schemes, employ coarse quantization coupled with oversampling. The standard mathematical model for the error accrued from such methods measures the performance of a given scheme by the rate at which the associated reconstruction error decays as a function of the oversampling ratio L > 1. It was recently shown that exponential accuracy of the form O(2(-r L)) can be achieved by appropriate one-bit Sigma Delta modulation schemes. However, the best known achievable rate constants r in this setting differ significantly from the general information theoretic lower bound. In this paper, we provide the first lower bound specific to coarse quantization, thus narrowing the gap between existing upper and lower bounds. In particular, our results imply a quantitative correspondence between the maximal signal amplitude and the best possible error decay rate. Our method draws from the theory of large deviations.

cs.IT↗
arXiv2010

On the complexity of Mumford-Shah type regularization, viewed as a relaxed sparsity constraint

We show that inverse problems with a truncated quadratic regularization are NP-hard in general to solve, or even approximate up to an additive error. This stands in contrast to the case corresponding to a finite-dimensional approximation to the Mumford-Shah functional, where the operator involved is the identity and for which polynomial-time solutions are known. Consequently, we confirm the infeasibility of any natural extension of the Mumford-Shah functional to general inverse problems. A connection between truncated quadratic minimization and sparsity-constrained minimization is also discussed.

math.OC↗
arXiv2009

Iterative thresholding meets free discontinuity problems

Free-discontinuity problems describe situations where the solution of interest is defined by a function and a lower dimensional set consisting of the discontinuities of the function. Hence, the derivative of the solution is assumed to be a `small' function almost everywhere except on sets where it concentrates as a singular measure. This is the case, for instance, in crack detection from fracture mechanics or in certain digital image segmentation problems. If we discretize such situations for numerical purposes, the free-discontinuity problem in the discrete setting can be re-formulated as that of finding a derivative vector with small components at all but a few entries that exceed a certain threshold. This problem is similar to those encountered in the field of `sparse recovery', where vectors with a small number of dominating components in absolute value are recovered from a few given linear measurements via the minimization of related energy functionals. Several iterative thresholding algorithms that intertwine gradient-type iterations with thresholding steps have been designed to recover sparse solutions in this setting. It is natural to wonder if and/or how such algorithms can be used towards solving discrete free-discontinuity problems. The current paper explores this connection, and, by establishing an iterative thresholding algorithm for discrete free-discontinuity problems, provides new insights on properties of minimizing solutions thereof.

math.NA↗
arXiv2008

Compressed Sensing with Cross Validation

Compressed Sensing decoding algorithms can efficiently recover an N dimensional real-valued vector x to within a factor of its best k-term approximation by taking m = 2klog(N/k) measurements y = Phi x. If the sparsity or approximate sparsity level of x were known, then this theoretical guarantee would imply quality assurance of the resulting compressed sensing estimate. However, because the underlying sparsity of the signal x is unknown, the quality of a compressed sensing estimate x* using m measurements is not assured. Nevertheless, we demonstrate that sharp bounds on the error || x - x* ||_2 can be achieved with almost no effort. More precisely, we assume that a maximum number of measurements m is pre-imposed; we reserve 4log(p) of the original m measurements and compute a sequence of possible estimates (x_j)_{j=1}^p to x from the m - 4log(p) remaining measurements; the errors ||x - x*_j ||_2 for j = 1, ..., p can then be bounded with high probability. As a consequence, numerical upper and lower bounds on the error between x and the best k-term approximation to x can be estimated for p values of k with almost no cost. Our observation has applications outside of compressed sensing as well.

math.NA↗