اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدMaximum Likelihood algorithm for parametric component separation in CMB experiments
115
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
We discuss an approach to the component separation of microwave, multi-frequency sky maps as those typically produced from Cosmic Microwave Background (CMB) Anisotropy data sets. The algorithm is based on the two step, parametric, likelihood-based technique recently elaborated on by Eriksen et al., (2006), where the foreground spectral parameters are estimated prior to the actual separation of the components. In contrast with the previous approaches, we accomplish the former task with help of an analytically-derived likelihood function for the spectral parameters, which, we show, yields estimates equal to the maximum likelihood values of the full multi-dimensional data problem. We then use these estimates to perform the second step via the standard, generalized-least-square-like procedure. We demonstrate that the proposed approach is equivalent to a direct maximization of the full data likelihood, which is recast in a computationally tractable form. We use the corresponding curvature matrices to characterize statistical properties of the recovered parameters. We incorporate in the formalism some of the essential features of the CMB data sets, such as inhomogeneous pixel domain noise, unknown map offsets as well as calibration errors and study their consequences for the separation. We find that the calibration is likely to have a dominant effect on the precision of the spectral parameter determination for a realistic CMB experiment. We apply the algorithm to simulated data and discuss the results. Our focus is on partial-sky, total-intensity and polarization, CMB experiments such as planned balloon-borne and ground-based efforts, however, the techniques presented here should be also applicable to the full-sky data as for instance, those produced by WMAP and anticipated from Planck.
قيم البحث
اقرأ أيضاً
A great deal of experimental effort is currently being devoted to the precise measurements of the cosmic microwave background (CMB) sky in temperature and polarisation. Satellites, balloon-borne, and ground-based experiments scrutinize the CMB sky at
multiple scales, and therefore enable to investigate not only the evolution of the early Universe, but also its late-time physics with unprecedented accuracy. The pipeline leading from time ordered data as collected by the instrument to the final product is highly structured. Moreover, it has also to provide accurate estimates of statistical and systematic uncertainties connected to the specific experiment. In this paper, we review likelihood approaches targeted to the analysis of the CMB signal at different scales, and to the estimation of key cosmological parameters. We consider methods that analyze the data in the spatial (i.e., pixel-based) or harmonic domain. We highlight the most relevant aspects of each approach and compare their performance.
We present in this paper the PolEMICA (Polarized Expectation-Maximization Independent Component Analysis) algorithm which is an extension to polarization of the SMICA (Spectral Matching Independent Component Analysis) temperature multi-detectors mult
i-components (MD-MC) component separation method (Delabrouille et al. 2003). This algorithm allows us to estimate blindly in harmonic space multiple physical components from multi-detectors polarized sky maps. Assuming a linear noisy mixture of components we are able to reconstruct jointly the anisotropies electromagnetic spectra of the components for each mode T, E and B, as well as the temperature and polarization spatial power spectra, TT, EE, BB, TE, TB and EB for each of the physical components and for the noise on each of the detectors. PolEMICA is specially developed to estimate the CMB temperature and polarization power spectra from sky observations including both CMB and foreground emissions. This has been tested intensively using as a first approach full sky simulations of the Planck satellite polarized channels for a 14-months nominal mission assuming a simplified linear sky model including CMB, and optionally Galactic synchrotron emission and a Gaussian dust emission. Finally, we have applied our algorithm to more realistic Planck full sky simulations, including synchrotron, realistic dust and free-free emissions.
The polarization modes of the cosmological microwave background are an invaluable source of information for cosmology, and a unique window to probe the energy scale of inflation. Extracting such information from microwave surveys requires disentangli
ng between foreground emissions and the cosmological signal, which boils down to solving a component separation problem. Component separation techniques have been widely studied for the recovery of CMB temperature anisotropies but quite rarely for the polarization modes. In this case, most component separation techniques make use of second-order statistics to discriminate between the various components. More recent methods, which rather emphasize on the sparsity of the components in the wavelet domain, have been shown to provide low-foreground, full-sky estimate of the CMB temperature anisotropies. Building on sparsity, the present paper introduces a new component separation technique dubbed PolGMCA (Polarized Generalized Morphological Component Analysis), which refines previous work to specifically tackle the estimation of the polarized CMB maps: i) it benefits from a recently introduced sparsity-based mechanism to cope with partially correlated components, ii) it builds upon estimator aggregation techniques to further yield a better noise contamination/non-Gaussian foreground residual trade-off. The PolGMCA algorithm is evaluated on simulations of full-sky polarized microwave sky simulations using the Planck Sky Model (PSM), which show that the proposed method achieve a precise recovery of the CMB map in polarization with low noise/foreground contamination residuals. It provides improvements with respect to standard methods, especially on the galactic center where estimating the CMB is challenging.
We revisit the problem of exact CMB likelihood and power spectrum estimation with the goal of minimizing computational cost through linear compression. This idea was originally proposed for CMB purposes by Tegmark et al. (1997), and here we develop i
t into a fully working computational framework for large-scale polarization analysis, adopting WMAP as a worked example. We compare five different linear bases (pixel space, harmonic space, noise covariance eigenvectors, signal-to-noise covariance eigenvectors and signal-plus-noise covariance eigenvectors) in terms of compression efficiency, and find that the computationally most efficient basis is the signal-to-noise eigenvector basis, which is closely related to the Karhunen-Loeve and Principal Component transforms, in agreement with previous suggestions. For this basis, the information in 6836 unmasked WMAP sky map pixels can be compressed into a smaller set of 3102 modes, with a maximum error increase of any single multipole of 3.8% at $ellle32$, and a maximum shift in the mean values of a joint distribution of an amplitude--tilt model of 0.006$sigma$. This compression reduces the computational cost of a single likelihood evaluation by a factor of 5, from 38 to 7.5 CPU seconds, and it also results in a more robust likelihood by implicitly regularizing nearly degenerate modes. Finally, we use the same compression framework to formulate a numerically stable and computationally efficient variation of the Quadratic Maximum Likelihood implementation that requires less than 3 GB of memory and 2 CPU minutes per iteration for $ell le 32$, rendering low-$ell$ QML CMB power spectrum analysis fully tractable on a standard laptop.
The log-concave maximum likelihood estimator (MLE) problem answers: for a set of points $X_1,...X_n in mathbb R^d$, which log-concave density maximizes their likelihood? We present a characterization of the log-concave MLE that leads to an algorithm
with runtime $poly(n,d, frac 1 epsilon,r)$ to compute a log-concave distribution whose log-likelihood is at most $epsilon$ less than that of the MLE, and $r$ is parameter of the problem that is bounded by the $ell_2$ norm of the vector of log-likelihoods the MLE evaluated at $X_1,...,X_n$.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
