Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userHow magnetic activity alters what we learn from stellar spectra
140
0
0.0
(
0
)
Ask ChatGPT about the research
No Arabic abstract
Magnetic fields and stellar spots can alter the equivalent widths of absorption lines in stellar spectra, varying during the activity cycle. This also influences the information that we derive through spectroscopic analysis. In this study we analyse high-resolution spectra of 211 Sun-like stars observed at different phases of their activity cycles, in order to investigate how stellar activity affects the spectroscopic determination of stellar parameters and chemical abundances. We observe that equivalent widths of lines can increase as a function of the activity index log R$^prime_{rm HK}$ during the stellar cycle, which also produces an artificial growth of the stellar microturbulence and a decrease in effective temperature and metallicity. This effect is visible for stars with activity indexes log R$^prime_{rm HK}$$geq$$-$5.0 (i.e., younger than 4-5 Gyr) and it is more significant at higher activity levels. These results have fundamental implications on several topics in astrophysics that are discussed in the paper, including stellar nucleosynthesis, chemical tagging, the study of Galactic chemical evolution, chemically anomalous stars, the structure of the Milky Way disk, stellar formation rates, photoevaporation of circumstellar disks, and planet hunting.
rate research
Read More
Nova SMC 2016 has been the most luminous nova known in the direction of the Magellanic Clouds. It turned into a very luminous supersoft X-ray source between day 16 and 28 after the optical maximum. We observed it with Chandra, the HRC-S camera and the Low Energy Transmission Grating (LETG) on 2016 November and 2017 January (days 39 and 88 after optical maximum), and with XMM-Newton on 2016 December (day 75). We detected the compact white dwarf (WD) spectrum as a luminous supersoft X-ray continuum with deep absorption features of carbon, nitrogen, magnesium, calcium, probably argon and sulfur on day 39, and oxygen, nitrogen and carbon on days 75 and 88. The spectral features attributed to the WD atmosphere are all blue-shifted, by about 1800 km/s on day 39 and up to 2100 km/s in the following observations. Spectral lines attributed to low ionization potential transitions in the interstellar medium are also observed. Assuming the distance of the Small Magellanic Cloud, the bolometric luminosity exceeded Eddington level for at least three months. A preliminary analysis with atmospheric models indicates effective temperature around 700,000 K on day 39, peaking at the later dates in the 850,000-900,000 K range, as expected for a 1.25 m(sol) WD. We suggest a possible classification as an oxygen-neon WD, but more precise modeling is needed to accurately determine the abundances. The X-ray light curves show large, aperiodic ux variability, not associated with spectral variability. We detected red noise, but did not find periodic or quasi-periodic modulations.
The learning rate is an information-theoretical quantity for bipartite Markov chains describing two coupled subsystems. It is defined as the rate at which transitions in the downstream subsystem tend to increase the mutual information between the two subsystems, and is bounded by the dissipation arising from these transitions. Its physical interpretation, however, is unclear, although it has been used as a metric for the sensing performance of the downstream subsystem. In this paper, we explore the behaviour of the learning rate for a number of simple model systems, establishing when and how its behaviour is distinct from the instantaneous mutual information between subsystems. In the simplest case, the two are almost equivalent. In more complex steady-state systems, the mutual information and the learning rate behave qualitatively distinctly, with the learning rate clearly now reflecting the rate at which the downstream system must update its information in response to changes in the upstream system. It is not clear whether this quantity is the most natural measure for sensor performance, and, indeed, we provide an example in which optimising the learning rate over a region of parameter space of the downstream system yields an apparently sub-optimal sensor.
We investigate the effects of multi-task learning using the recently introduced task of semantic tagging. We employ semantic tagging as an auxiliary task for three different NLP tasks: part-of-speech tagging, Universal Dependency parsing, and Natural Language Inference. We compare full neural network sharing, partial neural network sharing, and what we term the learning what to share setting where negative transfer between tasks is less likely. Our findings show considerable improvements for all tasks, particularly in the learning what to share setting, which shows consistent gains across all tasks.
We construct a simple and robust approach for deriving constraints on magnetic fields in galaxy clusters from rotation measure (RM) maps. Relaxing the commonly used assumptions of a correlation between the magnetic field strength and the plasma density and of a power-law (in wave number) magnetic field power spectrum, and using an efficient numerical analysis method, we test the consistency of a wide range of magnetic field models with RM maps of 11 extended sources in 5 clusters, for which the data were made available to us. We show that the data reveal no indication for a radial dependence of the average magnetic field strength, and in particular no indication for a correlation between the gas density and the field strength. The RM maps of a considerable fraction of the sources either require or are consistent with the presence of a spatially uniform magnetic field of a relatively small strength, 0.02-0.3 muG, which contributes significantly to the RM. The RM maps of all but one source do not require a power-law magnetic field power spectrum, and most are consistent with a power spectrum dominated by a single wave length. The uncertainties in the magnetic field strengths (and spatial correlation lengths) derived from RM maps exceed an order of magnitude (and often more). These uncertainties imply, in particular, that there is no indication in current RM data for a systematic difference between the magnetic field strengths in radio-halo clusters and in radio-quiet clusters. With the improvement expected in the near future of the quality and quantity of RM data, our analysis method will enable one to derive more accurate constraints on magnetic fields in galaxy clusters.
Learning problems form an important category of computational tasks that generalizes many of the computations researchers apply to large real-life data sets. We ask: what concept classes can be learned privately, namely, by an algorithm whose output does not depend too heavily on any one input or specific training example? More precisely, we investigate learning algorithms that satisfy differential privacy, a notion that provides strong confidentiality guarantees in contexts where aggregate information is released about a database containing sensitive information about individuals. We demonstrate that, ignoring computational constraints, it is possible to privately agnostically learn any concept class using a sample size approximately logarithmic in the cardinality of the concept class. Therefore, almost anything learnable is learnable privately: specifically, if a concept class is learnable by a (non-private) algorithm with polynomial sample complexity and output size, then it can be learned privately using a polynomial number of samples. We also present a computationally efficient private PAC learner for the class of parity functions. Local (or randomized response) algorithms are a practical class of private algorithms that have received extensive investigation. We provide a precise characterization of local private learning algorithms. We show that a concept class is learnable by a local algorithm if and only if it is learnable in the statistical query (SQ) model. Finally, we present a separation between the power of interactive and noninteractive local learning algorithms.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
