Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

The world is in the grip of the COVID-19 pandemic. Public health measures that can reduce the risk of infection and death in addition to quarantines are desperately needed. This article reviews the roles of vitamin D in reducing the risk of respiratory tract infections, knowledge about the epidemiology of influenza and COVID-19, and how vitamin D supplementation might be a useful measure to reduce risk. Through several mechanisms, vitamin D can reduce risk of infections. Those mechanisms include inducing cathelicidins and defensins that can lower viral replication rates and reducing concentrations of pro-inflammatory cytokines that produce the inflammation that injures the lining of the lungs, leading to pneumonia, as well as increasing concentrations of anti-inflammatory cytokines. Several observational studies and clinical trials reported that vitamin D supplementation reduced the risk of influenza, whereas others did not. Evidence supporting the role of vitamin D in reducing risk of COVID-19 includes that the outbreak occurred in winter, a time when 25-hydroxyvitamin D (25(OH)D) concentrations are lowest; that the number of cases in the Southern Hemisphere near the end of summer are low; that vitamin D deficiency has been found to contribute to acute respiratory distress syndrome; and that case-fatality rates increase with age and with chronic disease comorbidity, both of which are associated with lower 25(OH)D concentration. To reduce the risk of infection, it is recommended that people at risk of influenza and/or COVID-19 consider taking 10,000 IU/d of vitamin D3 for a few weeks to rapidly raise 25(OH)D concentrations, followed by 5000 IU/d. The goal should be to raise 25(OH)D concentrations above 40–60 ng/mL (100–150 nmol/L). For treatment of people who become infected with COVID-19, higher vitamin D3 doses might be useful. Randomized controlled trials and large population studies should be conducted to evaluate these recommendations.

Evidence that Vitamin D Supplementation Could Reduce Risk of Influenza and COVID-19 Infections and Deaths.

Our Galaxy, the Milky Way, is a benchmark for understanding disk galaxies. It is the only galaxy whose formation history can be studied using the full distribution of stars from faint dwarfs to supergiants. The oldest components provide us with unique insight into how galaxies form and evolve over billions of years. The Galaxy is a luminous (L⋆) barred spiral with a central box/peanut bulge, a dominant disk, and a diffuse stellar halo. Based on global properties, it falls in the sparsely populated “green valley” region of the galaxy color-magnitude diagram. Here we review the key integrated, structural and kinematic parameters of the Galaxy, and point to uncertainties as well as directions for future progress. Galactic studies will continue to play a fundamental role far into the future because there are measurements that can only be made in the near field and much of contemporary astrophysics depends on such observations.

The Galaxy in Context: Structural, Kinematic, and Integrated Properties

Aims. The aim of this paper is to explore and map the age and abundance structure of the stars in the nearby Galactic disk. Methods. We have conducted a high-resolution spectroscopic study of 714 F and G dwarf and subgiant stars in the Solar neighbourhood. The star sample has been kinematically selected to trace the Galactic thin and thick disks to their extremes, the metal-rich stellar halo, sub-structures in velocity space such as the Hercules stream and the Arcturus moving group, as well as stars that cannot (kinematically) be associated with either the thin disk or the thick disk. The determination of stellar parameters and elemental abundances is based on a standard analysis using equivalent widths and one-dimensional, plane-parallel model atmospheres calculated under the assumption of local thermodynamical equilibrium (LTE). The spectra have high resolution (R = 40 000-110 000) and high signal-to-noise)S/V = 150-300) and were obtained with the FEROS spectrograph on the ESO 1.5 in and 2.2 in telescopes, the SOFIN and PIES spectrographs on the Nordic Optical Telescope, the LIVES spectrograph on the E50 Very Large Telescope, the HARPS spectrograph on the ESO 3.6 m telescope, and the MIKE spectrograph on the Magellan Clay telescope. The abundances from individual Fe I lines were were corrected for non-LTE effects in every step of the analysis. Results. We present stellar parameters, stellar ages, kinematical parameters, orbital parameters, and detailed elemental abundances for 0, Na, Mg, Al, Si, Ca, Ti, Cr, Fe, Ni, Zn, Y. and Ba for 714 nearby 12 and G dwarf stars. Our data show that there is an old and a-enhanced disk population, and a younger and less a-enhanced disk population. While they overlap greatly in metallicity between 0.7 < [Fe/HI] less than or similar to +0.1, they show a bimodal distribution in [alpha/Fe]. This bimodality becomes even clearer if stars where stellar parameters and abundances show larger uncertainties (T-eff less than or similar to 5400 K) are discarded, showing that it is important to constrain the data set to a narrow range in the stellar parameters if small differences between stellar populations are to be revealed. In addition, we find that the a-enhanced population has orbital parameters placing the stellar birthplaces in the inner Galactic disk while the loss-alpha stars mainly come from the outer Galactic disk, fully consistent with the recent claims of a short scale-length for the alpha-enhanced Galactic thick disk. We have also investigated the properties of the Hercules stream and the Arcturus moving group and find that neither of them presents chemical or age signatures that could suggest that they are disrupted clusters or extragalactic accretion remnants from ancient merger events. Instead, they are most likely dynamical features originating within the Galaxy. We have also discovered that a standard 1D. LTE analysis, utilising ionisation and excitation balance of Fe I and Fen lines produces a flat lower main sequence. As the exact cause for this effect is unclear we chose to apply an empirical correction. Turn-off stars and more evolved stars appear to be unaffected. (Less)

/pdf/exploring-the-milky-way-stellar-disk-a-detailed-elemental-2cyhvnbv0r.pdf

Exploring the Milky Way stellar disk - A detailed elemental abundance study of 714 F and G dwarf stars in the solar neighbourhood

This work aims to present our current best physical understanding of common-envelope evolution (CEE). We highlight areas of consensus and disagreement, and stress ideas which should point the way forward for progress in this important but long-standing and largely unconquered problem. Unusually for CEE-related work, we mostly try to avoid relying on results from population synthesis or observations, in order to avoid potentially being misled by previous misunderstandings. As far as possible we debate all the relevant issues starting from physics alone, all the way from the evolution of the binary system immediately before CEE begins to the processes which might occur just after the ejection of the envelope. In particular, we include extensive discussion about the energy sources and sinks operating in CEE, and hence examine the foundations of the standard energy formalism. Special attention is also given to comparing the results of hydrodynamic simulations from different groups and to discussing the potential effect of initial conditions on the differences in the outcomes. We compare current numerical techniques for the problem of CEE and also whether more appropriate tools could and should be produced (including new formulations of computational hydrodynamics, and attempts to include 3D processes within 1D codes). Finally we explore new ways to link CEE with observations. We compare previous simulations of CEE to the recent outburst from V1309 Sco, and discuss to what extent post-common-envelope binaries and nebulae can provide information, e.g. from binary eccentricities, which is not currently being fully exploited.

/pdf/common-envelope-evolution-where-we-stand-and-how-we-can-move-1aki8lp6t7.pdf

Common envelope evolution: where we stand and how we can move forward

A ketogenic diet consumed during radiotherapy improves several aspects of quality of life and metabolic health in women with breast cancer.

/pdf/corticosteroids-compromise-survival-in-glioblastoma-in-part-55xqt3c5io.pdf

Corticosteroids compromise survival in glioblastoma in part through their elevation of blood glucose levels.

We investigate the performance of some common machine learning techniques in identifying blue horizontal branch (BHB) stars from photometric data. To train the machine learning algorithms, we use previously published spectroscopic identifications of BHB stars from Sloan digital sky survey (SDSS) data. We investigate the performance of three different techniques, namely k nearest neighbour classification, kernel density estimation for discriminant analysis and a support vector machine (SVM). We discuss the performance of the methods in terms of both completeness (what fraction of input BHB stars are successfully returned as BHB stars) and contamination (what fraction of contaminating sources end up in the output BHB sample). We discuss the prospect of trading off these values, achieving lower contamination at the expense of lower completeness, by adjusting probability thresholds for the classification. We also discuss the role of prior probabilities in the classification performance, and we assess via simulations the reliability of the dataset used for training. Overall it seems that no-prior gives the best completeness, but adopting a prior lowers the contamination. We find that the support vector machine generally delivers the lowest contamination for a given level of completeness, and so is our method of choice. Finally, we classify a large sample of SDSS Data Release 7 (DR7) photometry using the SVM trained on the spectroscopic sample. We identify 27 074 probable BHB stars out of a sample of 294 652 stars. We derive photometric parallaxes and demonstrate that our results are reasonable by comparing to known distances for a selection of globular clusters. We attach our classifications, including probabilities, as an electronic table, so that they can be used either directly as a BHB star catalogue, or as priors to a spectroscopic or other classification method. We also provide our final models so that they can be directly applied to new data.

/pdf/photometric-identification-of-blue-horizontal-branch-stars-4um7vb7tsc.pdf

Photometric identification of blue horizontal branch stars

Retracted: The Safety of COVID-19 Vaccinations-We Should Rethink the Policy

During the course of evolution, up until the agricultural revolution, environmental fluctuations forced the human species to develop a flexible metabolism in order to adapt its energy needs to various climate, seasonal and vegetation conditions. Metabolic flexibility safeguarded human survival independent of food availability. In modern times, humans switched their primal lifestyle towards a constant availability of energy-dense, yet often nutrient-deficient, foods, persistent psycho-emotional stressors and a lack of exercise. As a result, humans progressively gain metabolic disorders, such as the metabolic syndrome, type 2 diabetes, non-alcoholic fatty liver disease, certain types of cancer, cardiovascular disease and Alzheimer´s disease, wherever the sedentary lifestyle spreads in the world. For more than 2.5 million years, our capability to store fat for times of food shortage was an outstanding survival advantage. Nowadays, the same survival strategy in a completely altered surrounding is responsible for a constant accumulation of body fat. In this article, we argue that the metabolic epidemic is largely based on a deficit in metabolic flexibility. We hypothesize that the modern energetic inflexibility, typically displayed by symptoms of neuroglycopenia, can be reversed by re-cultivating suppressed metabolic programs, which became obsolete in an affluent environment, particularly the ability to easily switch to ketone body and fat oxidation. In a simplified model, the basic metabolic programs of humans' primal hunter-gatherer lifestyle are opposed to the current sedentary lifestyle. Those metabolic programs, which are chronically neglected in modern surroundings, are identified and conclusions for the prevention of chronic metabolic diseases are drawn.

Rainer J. Klement

Papers

A ketogenic diet consumed during radiotherapy improves several aspects of quality of life and metabolic health in women with breast cancer.

Corticosteroids compromise survival in glioblastoma in part through their elevation of blood glucose levels.

Photometric identification of blue horizontal branch stars

Retracted: The Safety of COVID-19 Vaccinations-We Should Rethink the Policy

The sedentary (r)evolution: Have we lost our metabolic flexibility? [version 1; referees: 2 approved]