اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدThe mass density field in simulated non-Gaussian scenarios
40
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
In this work we study the properties of the mass density field in the non-Gaussian world models simulated by Grossi et al. 2007. In particular we focus on the one-point density probability distribution function of the mass density field in non-Gausian models with quadratic non-linearities quantified by the usual parameter f_NL. We find that the imprints of primordial non-Gaussianity are well preserved in the negative tail of the probability function during the evolution of the density perturbation. The effect is already noticeable at redshifts as large as 4 and can be detected out to the present epoch. At z=0 we find that the fraction of the volume occupied by regions with underdensity delta < -0.9, typical of voids, is about 1.3 per cent in the Gaussian case and increases to ~2.2 per cent if f_NL=-1000 while decreases to ~0.5 per cent if f_NL=+1000. This result suggests that void-based statistics may provide a powerful method to detect non-Gaussianity even at low redshifts which is complementary to the measurements of the higher-order moments of the probability distribution function like the skewness or the kurtosis for which deviations from the Gaussian case are detected at the 25-50 per cent level.
قيم البحث
اقرأ أيضاً
We have performed high-resolution cosmological N-body simulations of a concordance LCDM model to study the evolution of virialized, dark matter haloes in the presence of primordial non-Gaussianity. Following a standard procedure, departures from Gaus
sianity are modeled through a quadratic Gaussian term in the primordial gravitational potential, characterized by a dimensionless non-linearity strength parameter f_NL. We find that the halo mass function and its redshift evolution closely follow the analytic predictions of Matarrese et al.(2000). The existence of precise analytic predictions makes the observation of rare, massive objects at large redshift an even more attractive test to detect primordial non-Gaussian features in the large scale structure of the universe.
Non-Gaussianity of the cosmological matter density field can be largely reduced by a local Gaussianization transformation (and its approximations such as the logrithmic transformation). Such behavior can be recasted as the Gaussian copula hypothesis,
and has been verified to very high accuracy at two-point level. On the other hand, statistically significant non-Gaussianities in the Gaussianized field have been detected in simulations. We point out that, this apparent inconsistency is caused by the very limited degrees of freedom in the copula function, which make it misleading as a diagnosis of residual non-Gaussianity in the Gaussianized field. Using the copula density, we highlight the departure from Gaussianity. We further quantify its impact in the predicted n-point correlation functions. We explore a remedy of the Gaussian copula hypothesis, which alleviates but not completely solves the above problems.
Autonomous vehicles are expected to navigate in complex traffic scenarios with multiple surrounding vehicles. The correlations between road users vary over time, the degree of which, in theory, could be infinitely large, thus posing a great challenge
in modeling and predicting the driving environment. In this paper, we propose a method to model multi-vehicle interactions using a stochastic vector field model and apply non-parametric Bayesian learning to extract the underlying motion patterns from a large quantity of naturalistic traffic data. We then use this model to reproduce the high-dimensional driving scenarios in a finitely tractable form. We use a Gaussian process to model multi-vehicle motion, and a Dirichlet process to assign each observation to a specific scenario. We verify the effectiveness of the proposed method on highway and intersection datasets from the NGSIM project, in which complex multi-vehicle interactions are prevalent. The results show that the proposed method can capture motion patterns from both settings, without imposing heroic prior, and hence demonstrate the potential application for a wide array of traffic situations. The proposed modeling method could enable simulation platforms and other testing methods designed for autonomous vehicle evaluation, to easily model and generate traffic scenarios emulating large scale driving data.
The abundance of collapsed objects in the universe, or halo mass function, is an important theoretical tool in studying the effects of primordially generated non-Gaussianities on the large scale structure. The non-Gaussian mass function has been calc
ulated by several authors in different ways, typically by exploiting the smallness of certain parameters which naturally appear in the calculation, to set up a perturbative expansion. We improve upon the existing results for the mass function by combining path integral methods and saddle point techniques (which have been separately applied in previous approaches). Additionally, we carefully account for the various scale dependent combinations of small parameters which appear. Some of these combinations in fact become of order unity for large mass scales and at high redshifts, and must therefore be treated non-perturbatively. Our approach allows us to do this, and to also account for multi-scale density correlations which appear in the calculation. We thus derive an accurate expression for the mass function which is based on approximations that are valid over a larger range of mass scales and redshifts than those of other authors. By tracking the terms ignored in the analysis, we estimate theoretical errors for our result and also for the results of others. We also discuss the complications introduced by the choice of smoothing filter function, which we take to be a top-hat in real space, and which leads to the dominant errors in our expression. Finally, we present a detailed comparison between the various expressions for the mass functions, exploring the accuracy and range of validity of each.
Various nonparametric approaches for Bayesian spectral density estimation of stationary time series have been suggested in the literature, mostly based on the Whittle likelihood approximation. A generalization of this approximation has been proposed
in Kirch et al. who prove posterior consistency for spectral density estimation in combination with the Bernstein-Dirichlet process prior for Gaussian time series. In this paper, we will extend the posterior consistency result to non-Gaussian time series by employing a general consistency theorem of Shalizi for dependent data and misspecified models. As a special case, posterior consistency for the spectral density under the Whittle likelihood as proposed by Choudhuri, Ghosal and Roy is also extended to non-Gaussian time series. Small sample properties of this approach are illustrated with several examples of non-Gaussian time series.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
