arXiv subjects

Laurence Park

Publications and source records attributed to Laurence Park.

4 recordsLinked to original sources

arXiv2025

EMU and the DRAGNs I: A Catalogue of DRAGNs

We present a catalogue of 3557 Double Radio sources associated with Active Galactic Nuclei (DRAGNs) from the First Pilot Survey of the Evolutionary Map of the Universe (EMU), observed at 944 MHz with the Australian Square Kilometre Array Pathfinder (ASKAP) telescope, covering 270 deg^2. We have extracted and identified each source by eye, tagged it with a morphological type and measured its parameters. The resulting catalogue will be used in subsequent papers to explore the properties of these sources, to train machine-learning algorithms for the detection of these sources in larger fields, and to compare with the results of Citizen Science projects, with the ultimate goal of understanding the physical processes that drive DRAGNs. Compared with earlier, lower sensitivity, catalogues, we find more diffuse structure and a plethora of more complex structures, ranging from wings of radio emission on the side of the jets, to types of object which have not been seen in earlier observations. As well as the well-known FR1 and FR2 sources, we find significant numbers of rare types of radio source such as Hybrid Morphology Radio Sources and one-sided jets, as well as a wide range of bent-tail and head-tail sources.

astro-ph.GA↗

arXiv2021

On the enumeration of integer tetrahedra

We consider the problem of enumerating integer tetrahedra of fixed perimeter (sum of side-lengths) and/or diameter (maximum side-length), up to congruence. As we will see, this problem is considerably more difficult than the corresponding problem for triangles, which has long been solved. We expect there are no closed-form solutions to the tetrahedron enumeration problems, but we explore the extent to which they can be approached via classical methods, such as orbit enumeration. We also discuss algorithms for computing the numbers, and present several tables and figures that can be used to visualise the data. Several intriguing patterns seem to emerge, leading to a number of natural conjectures. The central conjecture is that the number of integer tetrahedra of perimeter $n$, up to congruence, is asymptotic to $n^5/C$ for some constant $C\approx 229000$.

math.CO↗

arXiv2020

Accelerated Bayesian Optimization throughWeight-Prior Tuning

Bayesian optimization (BO) is a widely-used method for optimizing expensive (to evaluate) problems. At the core of most BO methods is the modeling of the objective function using a Gaussian Process (GP) whose covariance is selected from a set of standard covariance functions. From a weight-space view, this models the objective as a linear function in a feature space implied by the given covariance K, with an arbitrary Gaussian weight prior ${\bf w} \sim \mathcal{N} ({\bf 0}, {\bf I})$. In many practical applications there is data available that has a similar (covariance) structure to the objective, but which, having different form, cannot be used directly in standard transfer learning. In this paper we show how such auxiliary data may be used to construct a GP covariance corresponding to a more appropriate weight prior for the objective function. Building on this, we show that we may accelerate BO by modeling the objective function using this (learned) weight prior, which we demonstrate on both test functions and a practical application to short-polymer fibre manufacture.

stat.ML↗

arXiv2018

Covariance Function Pre-Training with m-Kernels for Accelerated Bayesian Optimisation

The paper presents a novel approach to direct covariance function learning for Bayesian optimisation, with particular emphasis on experimental design problems where an existing corpus of condensed knowledge is present. The method presented borrows techniques from reproducing kernel Banach space theory (specifically m-kernels) and leverages them to convert (or re-weight) existing covariance functions into new, problem-specific covariance functions. The key advantage of this approach is that rather than relying on the user to manually select (with some hyperparameter tuning and experimentation) an appropriate covariance function it constructs the covariance function to specifically match the problem at hand. The technique is demonstrated on two real-world problems - specifically alloy design and short-polymer fibre manufacturing - as well as a selected test function.

stat.ML↗