arXiv subjects
Publications and source records attributed to Mark Reid.
This document presents the draft specification for delivering machine learning services over HTTP, developed as part of the Protocols and Structures for Inference project, which concluded in 2013. It presents the motivation for providing machine learning as a service, followed by a description of the essential and optional components of such a service.
The Event Horizon Telescope (EHT), a global very long baseline interferometric array observing at a wavelength of 1.3 mm, detected the first image of the M87 supermassive black hole (SMBH). M87 is a giant elliptical galaxy at the center of Virgo cluster, which is expected to have formed through merging of cluster galaxies. Consequently M87* hosted mergers of black holes through dynamical friction and could have one or multiple binary companions with a low mass ratio at large separations. We show that a long-term monitoring of the M87 SMBH image over $\sim$1 year with absolute positional accuracy of 1$\approx\mu as$ could detect such binary companions and exclude a large parameter space in semi major axis ($a_0$) and mass ratio ($q$), which is currently not constrained. Moreover, the presence of the accretion disk around M87* excludes a binary companion with $a_0\approx$ of order a mili parsec, as otherwise the accretion disk would have been tidally disrupted.
Recent advances in VLBI have led to astrometric accuracy exceeding that of the Gaia mission goals. This chapter describes some important astrophysical problems that can be addressed with sub-milliarcsecond imaging and micro-arcsecond astrometry using the ngVLA with long baselines.
In combination with observations of the Cosmic Microwave Background, a measurement of the Hubble Constant provides a direct test of the standard $\Lambda$CDM cosmological model and a powerful constraint on the equation of state of dark energy. Observations of circumnuclear water vapor megamasers at 22 GHz in nearby active galaxies can be used to measure their distances, geometrically, and thereby provide a direct, one step measurement of the Hubble Constant. The measurement is independent of distance ladders and standard candles. With present-day instrumentation, the Megamaser Cosmology Project is expected to reach a $\sim$4% measurement using the megamaser technique, based on distances to fewer than 10 megamaser systems. A goal of the observational cosmology community is to reach a percent-level measurement of H$_0$ from several independent astrophysical measurements to minimize the systematics. The ngVLA will provide the sensitivity at 22 GHz required to discover and measure the additional megamaser disks that will enable an H$_0$ measurement at the $\sim$1% level.
The nature of the spiral structure of the Milky Way has long been debated. Only in the last decade have astronomers been able to accurately measure distances to a substantial number of high-mass star-forming regions, the classic tracers of spiral structure in galaxies. We report distance measurements at radio wavelengths using the Very Long Baseline Array for eight regions of massive star formation near the Local spiral arm of the Milky Way. Combined with previous measurements, these observations reveal that the Local Arm is larger than previously thought, and both its pitch angle and star formation rate are comparable to those of the Galaxy's major spiral arms, such as Sagittarius and Perseus. Toward the constellation Cygnus, sources in the Local Arm extend for a great distance along our line of sight and roughly along the solar orbit. Because of this orientation, these sources cluster both on the sky and in velocity to form the complex and long enigmatic Cygnus X region. We also identify a spur that branches between the Local and Sagittarius spiral arms.
This white paper discusses how a "next-generation" Very Large Array (ngVLA) operating in the frequency range 1-116 GHz could be a groundbreaking tool to study the detailed astrophysics of the "matter cycle" in the Milky Way and other galaxies. If optimized for high brightness sensitivity, the ngVLA would bring detailed microwave spectroscopy and modeling of the full radio spectral energy distribution into regular use as survey tools at resolutions of 0.1- 1 arcseconds. This wavelength range includes powerful diagnostics of density, excitation, and chemistry in the cold ISM, as well as multiple tracers of the rate of recent star formation, the magnetic field, shocks, and properties of the ionized ISM. We highlight design considerations that would make this facility revolutionary in this area, the foremost of which is a large amount of collecting area on ~km-length baselines. We also emphasize the strong case for harnessing the large proposed collecting area of the ngVLA for very long baseline applications as part of the core design. This would allow measurements of the three dimensional space motions of galaxies to beyond the Local Group and mapping of the Milky Way out to the far side of the disk. Finally, we discuss the gains from the proposed combination of very high resolution and sensitivity to thermal emission, which include observing the feeding of black holes and resolving forming protoclusters.
We discuss the unique opportunities for maser astrometry with the inclusion of the Square Kilometre Array (SKA) in Very Long Baseline Interferometry (VLBI) networks. The first phase of the SKA will enable observations of hydroxyl and methanol masers, positioning the latter to an accuracy of 5 microarcseconds, and the second phase may allow water maser observations. These observations will provide trigonometric distances with errors as small as 1%. The unrivalled sensitivity of the SKA will enable large-scale surveys and, through joint operations, will turn any VLBI network into a fast astrometry device. Both evolved stars and high mass star formation regions will be accessible throughout the (Southern) Milky Way, completing our understanding of the content, dynamics and history of our Galaxy. Maser velocities and proper motions will be measurable in the Local Group of galaxies and beyond, providing new insights into their kinematics and evolution.
Adding VLBI capability to the SKA arrays will greatly broaden the science of the SKA, and is feasible within the current specifications. SKA-VLBI can be initially implemented by providing phased-array outputs for SKA1-MID and SKA1-SUR and using these extremely sensitive stations with other radio telescopes, and in SKA2 by realising a distributed configuration providing baselines up to thousands of km, merging it with existing VLBI networks. The motivation for and the possible realization of SKA-VLBI is described in this paper.
We propose a novel hybrid loss for multiclass and structured prediction problems that is a convex combination of a log loss for Conditional Random Fields (CRFs) and a multiclass hinge loss for Support Vector Machines (SVMs). We provide a sufficient condition for when the hybrid loss is Fisher consistent for classification. This condition depends on a measure of dominance between labels--specifically, the gap between the probabilities of the best label and the second best label. We also prove Fisher consistency is necessary for parametric consistency when learning models such as CRFs. We demonstrate empirically that the hybrid loss typically performs least as well as--and often better than--both of its constituent losses on a variety of tasks, such as human action recognition. In doing so we also provide an empirical comparison of the efficacy of probabilistic and margin based approaches to multiclass and structured prediction.
We report a new geometric maser distance estimate to the active galaxy NGC 4258. The data for the new model are maser line-of-sight velocities and sky positions from 18 epochs of Very Long Baseline Interferometry observations, and line-of-sight accelerations measured from a 10-year monitoring program of the 22 GHz maser emission of NGC 4258. The new model includes both disk warping and confocal elliptical maser orbits with differential precession. The distance to NGC 4258 is 7.60 +/- 0.17 +/- 0.15 Mpc, a 3% uncertainty including formal fitting and systematic terms. The resulting Hubble Constant, based on the use of the Cepheid Variables in NGC 4258 to recalibrate the Cepheid distance scale (Riess et al. 2011), is H_0 = 72.0 +/- 3.0 km/s/Mpc.
We consider composite loss functions for multiclass prediction comprising a proper (i.e., Fisher-consistent) loss over probability distributions and an inverse link function. We establish conditions for their (strong) convexity and explore the implications. We also show how the separation of concerns afforded by using this composite representation allows for the design of families of losses with the same Bayes risk.
We show that the variational representations for f-divergences currently used in the literature can be tightened. This has implications to a number of methods recently proposed based on this representation. As an example application we use our tighter representation to derive a general f-divergence estimator based on two i.i.d. samples and derive the dual program for this estimator that performs well empirically. We also point out a connection between our estimator and MMD.
A long standing goal in astrophysics is to directly observe the immediate environment of a black hole with angular resolution comparable to the event horizon. Realizing this goal would open a new window on the study of General Relativity in the strong field regime, accretion and outflow processes at the edge of a black hole, the existence of an event horizon, and fundamental black hole physics (e.g., spin). Steady long-term progress on improving the capability of Very Long Baseline Interferometry (VLBI) at short wavelengths has now made it extremely likely that this goal will be achieved within the next decade. The most compelling evidence for this is the recent observation by 1.3mm VLBI of Schwarzschild radius scale structure in SgrA*, the compact source of radio, submm, NIR and xrays at the center of the Milky Way. SgrA* is thought to mark the position of a ~4 million solar mass black hole, and because of its proximity and estimated mass presents the largest apparent event horizon size of any black hole candidate in the Universe. Over the next decade, existing and planned mm/submm facilities will be combined into a high sensitivity, high angular resolution "Event Horizon Telescope" that will bring us as close to the edge of black hole as we will come for decades. This white paper describes the science case for mm/submm VLBI observations of both SgrA* and M87 (a radio loud AGN of a much more luminous class that SgrA*). We emphasize that while there is development and procurement involved, the technical path forward is clear, and the recent successful observations have removed much of the risk that would normally be associated with such an ambitious project.
A spectral-line survey of IRC+10216 in the 345 GHz band has been undertaken with the Submillimeter Array. Although not yet completed, it has already yielded a fairly large sample of narrow molecular emission lines with line-widths indicating expansion velocities of ~4 km/s, less than 3 times the well-known value of the terminal expansion velocity (14.5 km/s) of the outer envelope. Five of these narrow lines have now been identified as rotational transitions in vibrationally excited states of previously detected molecules: the v=1, J=17--16 and J=19--18 lines of Si34S and 29SiS and the v=2, J=7--6 line of CS. Maps of these lines show that the emission is confined to a region within ~60 AU of the star, indicating that the narrow-line emission is probing the region of dust-formation where the stellar wind is still being accelerated.
We present sensitive Very Large Array observations with an angular resolution of a few arcseconds of the $J= 1 - 0$ line of SiO in the $v$=1 and 2 vibrationally excited states toward a sample of 60 Galactic regions in which stars of high or intermediate mass are currently forming and/or have recently formed. We report the detection of SiO maser emission in \textit{both} vibrationally excited transitions toward only three very luminous regions: Orion-KL, W51N and Sgr B2(M). Toward all three, SiO maser emission had previously been reported, in Orion-KL in both lines, in W51N only in the $v=2$ line and in Sgr B2(M) only in the $v=1$ line. Our work confirms that SiO maser emission in star-forming regions is a rare phenomenon, indeed, that requires special, probably extreme, physical and chemical conditions not commonly found. In addition to this SiO maser survey, we also present images of the simultaneously observed 7 mm continuum emission from a subset of our sample of star-forming regions where such emission was detected. This is in most cases likely to be free-free emission from compact- and ultracompact-HII regions.
A wealth of new information about the structure of the maser disk in NGC 4258 has been obtained from a series of 18 VLBA observations spanning three years, as well as from 32 additional epochs of spectral monitoring data from 1994 to the present, acquired with the VLA, Effelsberg, and GBT. The warp of the disk has been defined precisely. The thickness of the maser disk has been measured to be 12 microarcseconds (FWHM), which is slightly smaller than previously quoted upper limits. Under the assumption that the masers trace the true vertical distribution of material in the disk, from the condition of hydrostatic equilibrium the sound speed is 1.5 km/s, corresponding to a thermal temperature of 600K. The accelerations of the high velocity maser components have been accurately measured for many features on both the blue and red side of the spectrum. The azimuthal offsets of these masers from the midline (the line through the disk in the plane of the sky) and derived projected offsets from the midline based on the warp model correspond well with the measured offsets. This result suggests that the masers are well described as discrete clumps of masing gas, which accurately trace the Keplerian motion of the disk. However, we have continued to search for evidence of apparent motions caused by ``phase effects.'' This work provides the foundation for refining the estimate of the distance to NGC 4258 through measurements of feature acceleration and proper motion. The refined estimate of this distance is expected to be announced in the near future.
The radio source Sgr A* at the center of our Galaxy is believed to be a 2.6 x 10^6 solar mass black hole which accretes gas from the winds of nearby stars. We show that limits on the X-ray and infrared emission from the Galactic Center provide an upper limit of ~ 8 x 10^{-5} solar masses per year on the mass accretion rate in Sgr A*. The advection-dominated accretion flow (ADAF) model favors a rate < 10^{-5} solar masses per year. In comparison, the Bondi accretion rate onto Sgr A*, estimated using the observed spatial distribution of mass losing stars and assuming non-interacting stellar winds, is ~ 3 x 10^{-5} solar masses per year. There is thus rough agreement between the Bondi, the ADAF, and the X-ray inferred accretion rates for Sgr A*. We discuss uncertainties in these estimates, emphasizing the importance of upcoming observations by the Chandra X-ray observatory (CXO) for tightening the X-ray derived limits.