arXiv subjects

S. C. Olhede

Publications and source records attributed to S. C. Olhede.

7 recordsLinked to original sources

arXiv2019

Statistical inference for network samples using subgraph counts

We consider that a network is an observation, and a collection of observed networks forms a sample. In this setting, we provide methods to test whether all observations in a network sample are drawn from a specified model. We achieve this by deriving, under the null of the graphon model, the joint asymptotic properties of average subgraph counts as the number of observed networks increases but the number of nodes in each network remains finite. In doing so, we do not require that each observed network contains the same number of nodes, or is drawn from the same distribution. Our results yield joint confidence regions for subgraph counts, and therefore methods for testing whether the observations in a network sample are drawn from: a specified distribution, a specified model, or from the same model as another network sample. We present simulation experiments and an illustrative example on a sample of brain networks where we find that highly creative individuals' brains present significantly more short cycles.

stat.ME↗

arXiv2017

Fractional Brownian motion, the Matern process, and stochastic modeling of turbulent dispersion

Stochastic process exhibiting power-law slopes in the frequency domain are frequently well modeled by fractional Brownian motion (fBm). In particular, the spectral slope at high frequencies is associated with the degree of small-scale roughness or fractal dimension. However, a broad class of real-world signals have a high-frequency slope, like fBm, but a plateau in the vicinity of zero frequency. This low-frequency plateau, it is shown, implies that the temporal integral of the process exhibits diffusive behavior, dispersing from its initial location at a constant rate. Such processes are not well modeled by fBm, which has a singularity at zero frequency corresponding to an unbounded rate of dispersion. A more appropriate stochastic model is a much lesser-known random process called the Matern process, which is shown herein to be a damped version of fractional Brownian motion. This article first provides a thorough introduction to fractional Brownian motion, then examines the details of the Matern process and its relationship to fBm. An algorithm for the simulation of the Matern process in O(N log N) operations is given. Unlike fBm, the Matern process is found to provide an excellent match to modeling velocities from particle trajectories in an application to two-dimensional fluid turbulence.

stat.ME↗

arXiv2017

Fast counting of medium-sized rooted subgraphs

We prove that counting copies of any graph $F$ in another graph $G$ can be achieved using basic matrix operations on the adjacency matrix of $G$. Moreover, the resulting algorithm is competitive for medium-sized $F$: our algorithm recovers the best known complexity for rooted 6-clique counting and improves on the best known for 9-cycle counting. Underpinning our proofs is the new result that, for a general class of graph operators, matrix operations are homomorphisms for operations on rooted graphs.

cs.DM↗

arXiv2009

Higher-Order Properties of Analytic Wavelets

The influence of higher-order wavelet properties on the analytic wavelet transform behavior is investigated, and wavelet functions offering advantageous performance are identified. This is accomplished through detailed investigation of the generalized Morse wavelets, a two-parameter family of exactly analytic continuous wavelets. The degree of time/frequency localization, the existence of a mapping between scale and frequency, and the bias involved in estimating properties of modulated oscillatory signals, are proposed as important considerations. Wavelet behavior is found to be strongly impacted by the degree of asymmetry of the wavelet in both the frequency and the time domain, as quantified by the third central moments. A particular subset of the generalized Morse wavelets, recognized as deriving from an inhomogeneous Airy function, emerge as having particularly desirable properties. These "Airy wavelets" substantially outperform the only approximately analytic Morlet wavelets for high time localization. Special cases of the generalized Morse wavelets are examined, revealing a broad range of behaviors which can be matched to the characteristics of a signal.

stat.ME↗

arXiv2006

Hyperanalytic Denoising

A new thresholding strategy for the estimation of a deterministic image immersed in noise is introduced. The threshold is combined with a wavelet decomposition, where the wavelet coefficient of the image at any fixed value of the decomposition index is estimated, via thresholding the observed coefficient depending on the value of both the magnitude of the observed coefficient as well as the magnitudes of coefficients of a set of additional images calculated from the observed image. The additional set of images is chosen so that the wavelet transforms of the full set of images have suitable deterministic and joint stochastic properties at a fixed scale and position index. Two different sets of additional images are suggested. The behaviour of the threshold criterion for a purely noisy image is investigated and a universal threshold is determined. The properties of the threshold for some typical deterministic signal structures are also given. The risk of an individual coefficient is determined, and calculated explicitly when the universal threshold is used, and some typical deterministic signal structures. The method is implemented on several examples and the theoretical risk reductions substantiated.

math.ST↗

arXiv2006

The Hyperanalytic Wavelet Transform

In this paper novel classes of 2-D vector-valued spatial domain wavelets are defined, and their properties given. The wavelets are 2-D generalizations of 1-D analytic wavelets, developed from the Generalized Cauchy-Riemann equations and represented as quaternionic functions. Higher dimensionality complicates the issue of analyticity, more than one `analytic' extension of a real function is possible, and an `analytic' analysis wavelet will not necessarily construct `analytic' decomposition coefficients. The decomposition of locally unidirectional and/or separable variation is investigated in detail, and two distinct families of hyperanalytic wavelet coefficients are introduced, the monogenic and the hypercomplex wavelet coefficients. The recasting of the analysis in a different frame of reference and its effect on the constructed coefficients is investigated, important issues for sampled transform coefficients. The magnitudes of the coefficients are shown to exhibit stability with respect to shifts in phase. Hyperanalytic 2-D wavelet coefficients enable the retrieval of a phase-and-magnitude description of an image in phase space, similarly to the description of a 1-D signal with the use of 1-D analytic wavelets, especially appropriate for oscillatory signals. Existing 2-D directional wavelet decompositions are related to the newly developed framework, and new classes of mother wavelets are introduced.

math.ST↗

arXiv2005

Multiple Multidimensional Morse Wavelets

We define a set of operators that localise a radial image in radial space and radial frequency simultaneously. We find the eigenfunctions of this operator and thus define a non-separable orthogonal set of radial wavelet functions that may be considered optimally concentrated over a region of radial space and radial scale space, defined via a doublet of parameters. We give analytic forms for their energy concentration over this region. We show how the radial function localisation operator can be generalised to an operator, localising any square integrable function in two dimensional Euclidean space. We show that the latter operator, with an appropriate choice of localisation region, approximately has the same eigenfunctions as the radial operator. Based on the radial wavelets we define a set of quaternionic valued wavelet functions that can extract local orientation for discontinuous signals and both orientation and phase structure for oscillatory signals. The full set of quaternionic wavelet functions are component wise orthogonal; hence their statistical properties are tractable, and we give forms for the variability of the estimates of the local phase and orientation, as well as the local energy of the image. By averaging estimates across wavelets, a substantial reduction in the variance is achieved.

math.ST↗