How to quantify fields or textures? A guide to the scattering transform

Cheng, Sihao; Ménard, Brice

doi:10.48550/arxiv.2112.01288

Cited by 12 publications

(22 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The combination of the above properties leads to a powerful estimator that can robustly capture the non-Gaussian information content encoded in a physical field, similar to a CNN, but while retaining the desired interpretability of conventional clustering statistics (e.g., correlation function) through a basis of a few well-defined and straightforward to extract WST coefficients. As opposed to the regular clustering statistics, in addition, the WST can extract higher-order correlations without raising the target field to high powers [67], which can cause instability. Last but not least, the WST has demonstrated the ability to better access the information content carried in physical fields with heavy-tailed probability distributions [67], a case that is particularly challenging for higher-order moments to describe [14].…”

Section: Wavelet Scattering Transformmentioning

confidence: 99%

“…As opposed to the regular clustering statistics, in addition, the WST can extract higher-order correlations without raising the target field to high powers [67], which can cause instability. Last but not least, the WST has demonstrated the ability to better access the information content carried in physical fields with heavy-tailed probability distributions [67], a case that is particularly challenging for higher-order moments to describe [14]. A pedagogical overview of various other properties of the WST (such as texture characterization or field generation) can be found in Ref.…”

Section: Wavelet Scattering Transformmentioning

confidence: 99%

“…A pedagogical overview of various other properties of the WST (such as texture characterization or field generation) can be found in Ref. [67], whereas the formal mathematical description is discussed in detail in Refs. [47,[53][54][55][56].…”

Section: Wavelet Scattering Transformmentioning

confidence: 99%

“…In light of the great promise held in the use of this estimator, the WST has been recently applied in the fields of astrophysics [57][58][59], cosmology [52,[60][61][62][63][64] and molecular chemistry [65,66] (a review can be found in Ref. [67]).…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Going Beyond the Galaxy Power Spectrum: an Analysis of BOSS Data with Wavelet Scattering Transforms

Valogiannis¹,

Dvorkin²

2022

Preprint

View full text Add to dashboard Cite

We perform the first application of the wavelet scattering transform (WST) on actual galaxy observations, through a WST analysis of the BOSS DR12 CMASS dataset. We lay out the detailed procedure on how to capture all necessary layers of realism for an application on data obtained from a spectroscopic survey, including the effects of redshift-space anisotropy, non-trivial survey geometry, the shortcomings of the dataset through a set of systematic weights and the Alcock-Paczynski distortion effect. In order to capture the cosmological dependence of the WST, we use galaxy mocks obtained from the state-of-the-art ABACUSSUMMIT simulations, tuned to match the anisotropic correlation function of the BOSS CMASS sample in the redshift range 0.46 < z < 0.60. Using our theory model for the WST coefficients, as well as for the first 2 multipoles of the galaxy power spectrum, that we use as reference, we perform a likelihood analysis of the CMASS data and obtain the posterior probability distributions of 4 cosmological parameters, {ω b , ωc, ns, σ8}, as well as the Hubble constant, derived from a fixed value of the angular size of the sound horizon at last scattering measured by the Planck satellite, all of which are marginalized over the 7 nuisance parameters of the Halo Occupation Distribution model. The WST is found to deliver a substantial improvement in the values of the predicted 1σ errors compared to the regular power spectrum, which are tighter by a factor in the range 3 − 6 in the case of flat and uninformative priors and by a factor of 4 − 28, when a Big Bang Nucleosynthesis prior is applied on the value of ω b . Furthermore, in the latter case, we obtain a 0.6% measurement of the Hubble constant. Our results are investigative and subject to certain approximations in our analysis, that we discuss in the text.

show abstract

Section: Wavelet Scattering Transformmentioning

confidence: 99%

Section: Wavelet Scattering Transformmentioning

confidence: 99%

Section: Wavelet Scattering Transformmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Going Beyond the Galaxy Power Spectrum: an Analysis of BOSS Data with Wavelet Scattering Transforms

Valogiannis¹,

Dvorkin²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Further, since they are pre-defined, the information is easier to interpret. For more in-depth discussions between these approaches we refer the reader to Cheng & Ménard (2021a).…”

Section: Introductionmentioning

confidence: 99%

Exploring the cosmic 21-cm signal from the Epoch of Reionisation using the Wavelet Scattering Transform

Greig,

Ting,

Kaurov

2022

Preprint

View full text Add to dashboard Cite

Detecting the cosmic 21-cm signal during the Epoch of Reionisation and Cosmic Dawn will reveal insights into the properties of the first galaxies and advance cosmological parameter estimation. Until recently, the primary focus for astrophysical parameter inference from the 21-cm signal centred on the power spectrum (PS). However, the cosmic 21-cm signal is highly non-Gaussian rendering the PS sub-optimal for characterising the cosmic signal. In this work, we introduce a new technique to analyse the non-Gaussian information in images of the 21-cm signal called the Wavelet Scattering Transform (WST). This approach closely mirrors that of convolutional neural networks with the added advantage of not requiring tuning or training of a neural network. Instead, it compresses the 2D spatial information into a set of coefficients making it easier to interpret while also providing a robust statistical description of the non-Gaussian information contained in the cosmic 21-cm signal. First, we explore the application of the WST to mock 21-cm images to gain valuable physical insights by comparing to the known behaviour from the 21-cm PS. Then we quantitatively explore the WST applied to the 21-cm signal by extracting astrophysical parameter constraints using Fisher Matrices from a realistic 1000 hr mock observation with the Square Kilometre Array. We find that: (i) the WST applied only to 2D images can outperform the 3D spherically averaged 21-cm PS, (ii) the excision of foreground contaminated modes can degrade the constraining power by a factor of ∼ 1.5 − 2 with the WST and (iii) higher cadences between the 21-cm images can further improve the constraining power.

show abstract

Ram pressure stripping in the z ∼ 0.5 galaxy cluster MS 0451.6-0305

Durret

Degott

Lobo

et al. 2022

A&A

View full text Add to dashboard Cite

Context. The pressure exerted by the ambient hot X-ray gas on cluster galaxies can lead to the presence of ram pressure stripped (RPS) galaxies, characterized by asymmetric shapes, and, in some cases, tails of blue stars and/or X-ray gas, with increased star formation. Aims. With the aim of increasing the known sample of RPS galaxies at redshift z ∼ 0.5, notably higher than most RPS samples presently known, we extended our searches for such galaxies to the cluster MS 0451.6-0305. Methods. Our study is based on Hubble Space Telescope (HST) imaging in the F814W band (corresponding to a restframe wavelength of 529 nm), covering a region of about 6 × 6 Mpc2, an eight magnitude ground-based catalogue with photometric redshifts, and a spectroscopic redshift catalogue. We defined as cluster members a spectroscopic redshift sample of 359 galaxies within ±4σv of the mean cluster velocity, and a photometric redshift sample covering the [0.48, 0.61] range. We searched for RPS galaxies in the HST images and tested the error on their classification with a Zooniverse collaboration. We also computed the phase space diagram of RPS candidates in the spectroscopic sample. Finally, we ran the LePhare stellar population synthesis code through the GAZPAR interface to analyze and compare the properties of RPS and non-RPS galaxies. Results. We find 56 and 273 RPS candidates in the spectroscopic and photometric redshift samples, respectively. They are distributed throughout the cluster and tend to avoid high density regions. The phase space diagram gives the percentages of virialized, backsplash, and infall galaxies. RPS galaxy candidates typically show rather high star formation rates, young ages, and relatively low masses. We compare all our results to those previously obtained for the massive merging cluster MACS J0717+3745, at a similar redshift. Conclusions. This study increases by at least 56 objects if we only consider galaxies with spectroscopic redshifts, and probably much more if galaxies with photometric redshifts are taken into account. This study increases the number of RPS candidates at redshift z > 0.5, and confirms that they host, on average, younger stellar populations and strongly form stars when compared with non-RPS counterparts. The fact that RPS candidates with spectroscopic and with photometric redshifts have comparable properties shows that large samples of such objects could be gathered based on multi-band photometry only, a promising result in view of the very large imaging surveys planned in the coming years (DES, Euclid, LSST, etc.).

show abstract

How to quantify fields or textures? A guide to the scattering transform

Cited by 12 publications

References 25 publications

Going Beyond the Galaxy Power Spectrum: an Analysis of BOSS Data with Wavelet Scattering Transforms

Going Beyond the Galaxy Power Spectrum: an Analysis of BOSS Data with Wavelet Scattering Transforms

Exploring the cosmic 21-cm signal from the Epoch of Reionisation using the Wavelet Scattering Transform

Ram pressure stripping in the z ∼ 0.5 galaxy cluster MS 0451.6-0305

Contact Info

Product

Resources

About