Related papers: Correcting for interloper contamination in the pow…
Given data with label noise (i.e., incorrect data), deep neural networks would gradually memorize the label noise and impair model performance. To relieve this issue, curriculum learning is proposed to improve model performance and…
We present the strategy used to identify and mitigate potential sources of angular systematics in the \textit{Euclid} spectroscopic galaxy survey, and we quantify their impact on galaxy clustering measurements and cosmological parameter…
Over the next few years new spectroscopic surveys (from the optical surveys of the Sloan Digital Sky Survey and the 2 degree Field survey through to space-based ultraviolet satellites such as GALEX) will provide the opportunity and…
We model a 21 cm intensity mapping survey in the redshift range 0.01<z<1.5 designed to simulate the skies as seen by future radio telescopes such as the Square Kilometre Array (SKA), including instrumental noise and Galactic foregrounds. In…
Due to their weak inductive bias, Multi-Layer Perceptrons (MLPs) have subpar performance at low-compute levels compared to standard architectures such as convolution-based networks (CNN). Recent work, however, has shown that the performance…
We present a joint likelihood analysis of the real-space power spectrum and bispectrum measured from a variety of halo and galaxy mock catalogs. A novel aspect of this work is the inclusion of nonlinear triangle configurations for the…
We explore the cosmological multitracer synergies between an emission line galaxy distribution from the Dark Energy Spectroscopic Instrument and a Tianlai Project 21-cm intensity map. We use simulated maps generated from a particle…
Learning with Noisy Labels (LNL) has attracted significant attention from the research community. Many recent LNL methods rely on the assumption that clean samples tend to have "small loss". However, this assumption always fails to…
Despite being robust to small amounts of label noise, convolutional neural networks trained with stochastic gradient methods have been shown to easily fit random labels. When there are a mixture of correct and mislabelled targets, networks…
The forthcoming generation of galaxy redshift surveys will sample the large-scale structure of the Universe over unprecedented volumes with high-density tracers. This advancement will make robust measurements of three-point clustering…
Spectral line intensity mapping has been proposed as a promising tool to efficiently probe the cosmic reionization and the large-scale structure. Without detecting individual sources, line intensity mapping makes use of all available…
We consider the problem of estimating a signal subspace in the presence of interference that contaminates some proportion of the received observations. Our emphasis is on detecting the contaminated observations so that the signal subspace…
We construct a simple but self-consistent analytic ionization model for rapid exploration of 21cm power spectrum observables in redshift space. It is fully described by the average ionization fraction $x_e(z)$ and HII patch size $R(z)$ and…
The information content of the minimum spanning tree (MST), used to capture higher-order statistics and information from the cosmic web, is compared to that of the power spectrum for a $\nu\Lambda$CDM model. The measurements are made in…
We study the impact of different bias and redshift-space models on the halo power spectrum, quantifying their effect by comparing the fit to a subset of realizations taken from the WizCOLA suite. These provide simulated power spectrum…
Falsely annotated samples, also known as noisy labels, can significantly harm the performance of deep learning models. Two main approaches for learning with noisy labels are global noise estimation and data filtering. Global noise…
Data analysis from upcoming large galaxy redshift surveys, such as Euclid and DESI will significantly improve constraints on cosmological parameters. To optimally extract the information from these galaxy surveys, it is important to control…
The next generation of galaxy surveys has the potential to substantially deepen our understanding of the Universe. This potential hinges on our ability to rigorously address systematic uncertainties. Until now, diagnosing systematic effects…
In this paper, we study the problem of learning image classification models with label noise. Existing approaches depending on human supervision are generally not scalable as manually identifying correct or incorrect labels is…
Marked power spectra provide a computationally efficient way to extract non-Gaussian information from the matter density field using the usual analysis tools developed for the power spectrum without the need for explicit calculation of…