Related papers: Differentiable Stochastic Halo Occupation Distribu…
We propose Deep Hierarchical Machine (DHM), a model inspired from the divide-and-conquer strategy while emphasizing representation learning ability and flexibility. A stochastic routing framework as used by recent deep neural…
When constructing galaxy mock catalogs based on suits of dark matter halo catalogs generated with approximated, calibrated or machine-learning approaches, the assignment of intrinsic properties for such tracers is a step of paramount…
We present a new suite of mock galaxy catalogs mimicking the low-redshift Universe, based on an updated halo occupation distribution (HOD) model and a scaling relation between optical properties and the neutral hydrogen (HI) content of…
Hashing has attracted increasing research attentions in recent years due to its high efficiency of computation and storage in image retrieval. Recent works have demonstrated the superiority of simultaneous feature representations and hash…
Effective field theory (EFT)-based full-shape analysis with simulation-based priors (SBPs) is a novel approach to galaxy clustering data analysis, which significantly boosts the constraining power by efficiently incorporating field-level…
Hamiltonian Monte Carlo (HMC) is a powerful and accurate method to sample from the posterior distribution in Bayesian inference. However, HMC techniques are computationally demanding for Bayesian neural networks due to the high…
Generating mocks for future sky surveys requires large volumes and high resolutions, which is computationally expensive even for fast simulations. In this work we try to develop numerical schemes to calibrate various halo and matter…
We use the projected correlation function w_p(r_p) of a volume-limited subsample of the Sloan Digital Sky Survey (SDSS) main galaxy redshift catalogue to measure the halo occupation distribution (HOD) of the galaxies of the sample.…
We present a set of mock redshift catalogs, constructed from N-body simulations, designed to mimic the DEEP2 survey. Galaxies with a range of luminosities are placed within virialized halos in the simulation using a variant of the halo…
We introduce the Hamming Ball Sampler, a novel Markov Chain Monte Carlo algorithm, for efficient inference in statistical models involving high-dimensional discrete state spaces. The sampling scheme uses an auxiliary variable construction…
Halo Occupation Distribution (HOD) is a model giving the average number of galaxies in a dark matter halo, function of its mass and other intrinsic properties, like distance from halo center, luminosity and redshift of its constituting…
Numerical simulations show that redshift space distortions (RSD) introduce strong scale dependence in the power spectra of halos, with ten percent deviations relative to linear theory predictions even on relatively large scales (k<0.1h/Mpc)…
Online hashing has attracted extensive research attention when facing streaming data. Most online hashing methods, learning binary codes based on pairwise similarities of training instances, fail to capture the semantic relationship, and…
Modelers use automatic differentiation (AD) of computation graphs to implement complex Deep Learning models without defining gradient computations. Stochastic AD extends AD to stochastic computation graphs with sampling steps, which arise…
A very large dynamic range with simultaneous capture of both large- and small-scales in the simulations of cosmic structures is required for correct modelling of many cosmological phenomena, particularly at high redshift. This is not always…
Robust Bayesian inference using density power divergence (DPD) has emerged as a promising approach for handling outliers in statistical estimation. Although the DPD-based posterior offers theoretical guarantees of robustness, its practical…
Full ray-tracing maps of gravitational lensing, constructed from N-Body simulations, represent a fundamental tool to interpret present and future weak lensing data. However the limitation of computational resources and storage capabilities…
We present a class of diffusion-based algorithms to draw samples from high-dimensional probability distributions given their unnormalized densities. Ideally, our methods can transport samples from a Gaussian distribution to a specified…
We analyze the halo occupation distribution (HOD), the probability for a halo of mass M to host a number of subhalos N, and two-point correlation function of galaxy-size dark matter halos using high-resolution dissipationless simulations of…
We study reinforcement learning in hybrid discrete-continuous action spaces, such as settings where the discrete component selects a regime (or index) and the continuous component optimizes within it -- a structure common in robotics,…