The huge amount of textual data on the Web has grown in the last few years rapidly creating unique contents of massive dimension. In a decision making context, one of the most relevant tasks is ...polarity classification of a text source, which is usually performed through supervised learning methods. Most of the existing approaches select the best classification model leading to over-confident decisions that do not take into account the inherent uncertainty of the natural language. In this paper, we pursue the paradigm of ensemble learning to reduce the noise sensitivity related to language ambiguity and therefore to provide a more accurate prediction of polarity. The proposed ensemble method is based on Bayesian Model Averaging, where both uncertainty and reliability of each single model are taken into account. We address the classifier selection problem by proposing a greedy approach that evaluates the contribution of each model with respect to the ensemble. Experimental results on gold standard datasets show that the proposed approach outperforms both traditional classification and ensemble methods.
•A novel ensemble learning methodology is proposed for polarity classification task.•A selection strategy is studied to reduce the search space of candidate ensembles.•The proposed model has been shown to be effective and efficient in several domains.
The Atacama Large Millimeter Array (ALMA) Large Program to INvestigate CII at Early times (ALPINE) targets the CII 158
μ
m line and the far-infrared continuum in 118 spectroscopically confirmed ...star-forming galaxies between
z
= 4.4 and
z
= 5.9. It represents the first large CII statistical sample built in this redshift range. We present details regarding the data processing and the construction of the catalogs. We detected 23 of our targets in the continuum. To derive accurate infrared luminosities and obscured star formation rates (SFRs), we measured the conversion factor from the ALMA 158
μ
m rest-frame dust continuum luminosity to the total infrared luminosity (
L
IR
) after constraining the dust spectral energy distribution by stacking a photometric sample similar to ALPINE in ancillary single-dish far-infrared data. We found that our continuum detections have a median
L
IR
of 4.4 × 10
11
L
⊙
. We also detected 57 additional continuum sources in our ALMA pointings. They are at a lower redshift than the ALPINE targets, with a mean photometric redshift of 2.5 ± 0.2. We measured the 850
μ
m number counts between 0.35 and 3.5 mJy, thus improving the current interferometric constraints in this flux density range. We found a slope break in the number counts around 3 mJy with a shallower slope below this value. More than 40% of the cosmic infrared background is emitted by sources brighter than 0.35 mJy. Finally, we detected the CII line in 75 of our targets. Their median CII luminosity is 4.8 × 10
8
L
⊙
and their median full width at half maximum is 252 km s
−1
. After measuring the mean obscured SFR in various CII luminosity bins by stacking ALPINE continuum data, we find a good agreement between our data and the local and predicted SFR–
L
CII
relations.
The survival of dust grains in galaxies depends on various processes. Dust can be produced in stars, it can grow in the interstellar medium and be destroyed by astration and interstellar shocks. In ...this paper, we assemble a few data samples of local and distant star-forming galaxies to analyse various dust-related quantities in low- and high-redshift galaxies, and to study how the relations linking the dust mass to the stellar mass and star formation rate evolve with redshift. We interpret the available data by means of chemical evolution models for discs and proto-spheroid (PSPH) starburst galaxies. In particular, we focus on the dust-to-stellar mass (DTS) ratio, as this quantity represents a true measure of how much dust per unit stellar mass survives the various destruction processes in galaxies and is observable. The theoretical models outline the strong dependence of this quantity on the underlying star formation history. Spiral galaxies are characterized by a nearly constant DTS as a function of the stellar mass and cosmic time, whereas PSPHs present an early steep increase of the DTS, which stops at a maximal value and decreases in the latest stages. In their late starburst phase, these models show a decrease of the DTS with their mass, which allows us to explain the observed anti-correlation between the DTS and the stellar mass. The observed redshift evolution of the DTS ratio shows an increase from z ~ 0 to z ~ 1, followed by a roughly constant behaviour at ... Our models indicate a steep decrease of the global DTS at early times, which implies an expected decrease of the DTS at larger redshift. (ProQuest: ... denotes formulae/symbols omitted.)
We present the ancillary data and basic physical measurements for the galaxies in the ALMA Large Program to Investigate C+ at Early Times (ALPINE) survey-the first large multiwavelength survey that ...aims at characterizing the gas and dust properties of 118 main-sequence galaxies at redshifts 4.4 < z < 5.9 via the measurement of emission at (64% at >3.5 ) and the surrounding far-infrared continuum in conjunction with a wealth of optical and near-infrared data. We outline in detail the spectroscopic data and selection of the galaxies as well as the ground- and space-based imaging products. In addition, we provide several basic measurements including stellar masses, star formation rates (SFR), rest-frame ultra-violet (UV) luminosities, UV continuum slopes (β), and absorption line redshifts, as well as H emission derived from Spitzer colors. We find that the ALPINE sample is representative of the 4 < z < 6 galaxy population selected by photometric methods and only slightly biased toward bluer colors (Δβ ∼ 0.2). Using as tracer of the systemic redshift (confirmed for one galaxy at z = 4.5 out of 118 for which we obtained optical λ3727 emission), we confirm redshifted Ly emission and blueshifted absorption lines similar to findings at lower redshifts. By stacking the rest-frame UV spectra in the rest frame, we find that the absorption lines in galaxies with high specific SFR are more blueshifted, which could be indicative of stronger winds and outflows.
We study the efficiency of galactic feedback in the early Universe by stacking the C II 158
μ
m emission in a large sample of normal star-forming galaxies at 4 <
z
< 6 from the ALMA Large Program ...to INvestigate C II at Early times (ALPINE) survey. Searching for typical signatures of outflows in the high-velocity tails of the stacked C II profile, we observe (i) deviations from a single-component Gaussian model in the combined residuals and (ii) broad emission in the stacked C II spectrum, with velocities of |
v
|≲500 km s
−1
. The significance of these features increases when stacking the subset of galaxies with star formation rates (SFRs) higher than the median (SFR
med
= 25
M
⊙
yr
−1
), thus confirming their star-formation-driven nature. The estimated mass outflow rates are comparable to the SFRs, yielding mass-loading factors of the order of unity (similarly to local star-forming galaxies), suggesting that star-formation-driven feedback may play a lesser role in quenching galaxies at
z
> 4. From the stacking analysis of the datacubes, we find that the combined C II core emission (|
v
|< 200 km s
−1
) of the higher-SFR galaxies is extended on physical sizes of ∼30 kpc (diameter scale), well beyond the analogous C II core emission of lower-SFR galaxies and the stacked far-infrared continuum. The detection of such extended metal-enriched gas, likely tracing circumgalactic gas enriched by past outflows, corroborates previous similar studies, confirming that baryon cycle and gas exchanges with the circumgalactic medium are at work in normal star-forming galaxies already at early epochs.
The dust mass function from z ∼0 to z ∼2.5 Pozzi, F; Calura, F; Zamorani, G ...
Monthly notices of the Royal Astronomical Society,
02/2020, Letnik:
491, Številka:
4
Journal Article
Recenzirano
Odprti dostop
ABSTRACT
We derive for the first time the dust mass function (DMF) in a wide redshift range, from z ∼ 0.2 up to z ∼ 2.5. In order to trace the dust emission, we start from a far-IR (160-μm) Herschel ...selected catalogue in the COSMOS field. We estimate the dust masses by fitting the far-IR data (λrest$\,\, \buildrel\gt \over \sim \,\,$50 μm) with a modified black body function and we present a detailed analysis to take into account the incompleteness in dust masses from a far-IR perspective. By parametrizing the observed DMF with a Schechter function in the redshift range 0.1 < z ≤ 0.25, where we are able to sample faint dust masses, we measure a steep slope (α ∼1.48), as found by the majority of works in the Local Universe. We detect a strong dust mass evolution, with $M_{\rm d}^{\star }$ at z ∼ 2.5 almost 1 dex larger than in the local Universe, combined with a decrease in their number density. Integrating our DMFs, we estimate the dust mass density (DMD), finding a broad peak at z ∼ 1, with a decrease by a factor of ∼ 3 towards z ∼ 0 and z ∼ 2.5. In general, the trend found for the DMD mostly agrees with the derivation of Driver et al., another DMD determination based also on far-IR detections, and with other measures based on indirect tracers.
We study the incidence of nuclear obscuration on a complete sample of 1310 active galactic nuclei (AGN) selected on the basis of their rest-frame 2-10 keV X-ray flux from the XMM-COSMOS survey, in ...the redshift range 0.3 < z < 3.5. We classify the AGN as obscured or unobscured on the basis of either the optical spectral properties and the overall SED or the shape of the X-ray spectrum. The two classifications agree in about 70 per cent of the objects, and the remaining 30 per cent can be further subdivided into two distinct classes: at low luminosities X-ray unobscured AGN do not always show signs of broad lines or blue/UV continuum emission in their optical spectra, most likely due to galaxy dilution effects; at high-luminosities broad-line AGN may have absorbed X-ray spectra, which hints at an increased incidence of small-scale (sub-parsec) dust-free obscuration. We confirm that the fraction of obscured AGN is a decreasing function of the intrinsic X-ray luminosity, while the incidence of absorption shows significant evolution only for the most luminous AGN, which appear to be more commonly obscured at higher redshift. We find no significant difference between the mean stellar masses and star formation rates of obscured and unobscured AGN hosts. We conclude that the physical state of the medium responsible for obscuration in AGN is complex and mainly determined by the radiation environment (nuclear luminosity) in a small region enclosed within the gravitational sphere of influence of the central black hole, but is largely insensitive to the wider scale galactic conditions.
ABSTRACT
We study the impact of active galactic nuclei (AGNs) on the CO spectral line energy distribution (SLED) of high-z galaxies. In particular, we want to assess if the CO SLED can be used as a ...probe of AGN activity. To this purpose, we develop a semi-analytical model that takes into account the radiative transfer and the clumpy structure of giant molecular clouds where the CO lines are excited, their distribution in the galaxy disc, and the torus obscuration of the AGN radiation. We study the joint effect on the CO SLED excitation of (i) the X-ray luminosity from the AGNs (LX), (ii) the size of the molecular disc, (iii) the inclination angle (Ω) of the torus with respect to the molecular disc, and (iv) the Giant Molecular Cloud (GMC) mean density. We also discuss the possibility of an enhanced cosmic ray ionization rate (CRIR). We find that the X-ray-dominated region (XDR) generated by the AGNs in every case enhances the CO SLED for J > 5, with increasing excitation of high-J CO lines for increasing X-ray luminosity. Because high-z galaxies are compact, the XDR typically encloses the whole disc, thus its effect can be more important with respect to lower redshift objects. The impact of the XDR can be disentangled from an enhanced CRIR either if $L_\mathrm{ X}\,\gt\, 10^{44} \rm \, erg\, s^{-1}$, or if Ω ≥ 60°. We finally provide predictions on the CO(7–6)/C ii and CO(17–16)/C ii ratios as a function of LX, which can be relevant for ALMA follow-up of galaxies and quasars previously detected in C ii.
The CERN-EU high-energy Reference Field (CERF) facility provides an almost unique high-energy workplace radiation field for calibrating and testing radiation protection instrumentation employed at ...high-energy accelerators and for aircraft and space dosimetry. For the available exposure locations, this work provides the particle fluences of the various components of the radiation field, and a new set of reference values for the currently used operational quantities, i.e. ambient dose equivalent, H*(10), and personal dose equivalent Hp(10), as well as for the newly proposed ICRU/ICRP quantities ambient dose, H*, and personal dose, Hp, that may be adopted in the future for radiation protection. Particle spectra and radiological quantities were computed with the latest versions of the FLUKA code and benchmarked against experimental data taken with one of the reference instruments used at the facility.
FLUKA is a general purpose Monte Carlo code able to describe the transport and interaction of any particle and nucleus type in complex geometries over an energy range extending from thermal neutrons ...to ultrarelativistic hadron collisions. It has many different applications in accelerator design, detector studies, dosimetry, radiation protection, medical physics, and space research. In 2019, CERN and INFN, as FLUKA copyright holders, together decided to end their formal collaboration framework, allowing them henceforth to pursue different pathways aimed at meeting the evolving requirements of the FLUKA user community, and at ensuring the long term sustainability of the code. To this end, CERN set up the FLUKA.CERN Collaboration
1
. This paper illustrates the physics processes that have been newly released or are currently implemented in the code distributed by the FLUKA.CERN Collaboration
2
under new licensing conditions that are meant to further facilitate access to the code, as well as intercomparisons. The description of coherent effects experienced by high energy hadron beams in crystal devices, relevant to promising beam manipulation techniques, and the charged particle tracking in vacuum regions subject to an electric field, overcoming a former lack, have already been made available to the users. Other features, namely the different kinds of low energy deuteron interactions as well as the synchrotron radiation emission in the course of charged particle transport in vacuum regions subject to magnetic fields, are currently undergoing systematic testing and benchmarking prior to release. FLUKA is widely used to evaluate radiobiological effects, with the powerful support of the Flair graphical interface, whose new generation (Available at
http://flair.cern
) offers now additional capabilities, e.g., advanced 3D visualization with photorealistic rendering and support for industry-standard volume visualization of medical phantoms. FLUKA has also been playing an extensive role in the characterization of radiation environments in which electronics operate. In parallel, it has been used to evaluate the response of electronics to a variety of conditions not included in radiation testing guidelines and standards for space and accelerators, and not accessible through conventional ground level testing. Instructive results have been obtained from Single Event Effects (SEE) simulations and benchmarks, when possible, for various radiation types and energies. The code has reached a high level of maturity, from which the FLUKA.CERN Collaboration is planning a substantial evolution of its present architecture. Moving towards a modern programming language allows to overcome fundamental constraints that limited development options. Our long term goal, in addition to improving and extending its physics performances with even more rigorous scientific oversight, is to modernize its structure to integrate independent contributions more easily and to formalize quality assurance through state-of-the-art software deployment techniques. This includes a continuous integration pipeline to automatically validate the codebase as well as automatic processing and analysis of a tailored physics-case test suite. With regard to the aforementioned objectives, several paths are currently envisaged, like finding synergies with Geant4, both at the core structure and interface level, this way offering the user the possibility to run with the same input different Monte Carlo codes and crosscheck the results.