• ### Towards Automatic Construction of Diverse, High-quality Image Dataset(1708.06495)

March 1, 2019 cs.CV, cs.MM
The availability of labeled image datasets has been shown critical for high-level image understanding, which continuously drives the progress of feature designing and models developing. However, constructing labeled image datasets is laborious and monotonous. To eliminate manual annotation, in this work, we propose a novel image dataset construction framework by employing multiple textual queries. We aim at collecting diverse and accurate images for given queries from the Web. Specifically, we formulate noisy textual queries removing and noisy images filtering as a multi-view and multi-instance learning problem separately. Our proposed approach not only improves the accuracy but also enhances the diversity of the selected images. To verify the effectiveness of our proposed approach, we construct an image dataset with 100 categories. The experiments show significant performance gains by using the generated data of our approach on several tasks, such as image classification, cross-dataset generalization, and object detection. The proposed method also consistently outperforms existing weakly supervised and web-supervised approaches.
• ### Stability and boundedness in AdS/CFT with double trace deformations(1709.00445)

Feb. 14, 2019 hep-th
Scalar fields on the bulk side of AdS/CFT correspondence can be assigned unconventional boundary conditions, related to the conventional one by Legendre transform. One can further perform double trace deformations which relate the two boundary conditions via renormalization group flow. Thinking of these operators as $S$ and $T$ transformations, respectively, we explore the $SL(2,{\bf R})$ family of models which naively emerges from repeatedly applying these operations. Depending on the parameters, the effective masses vary and can render the theory unstable. However, unlike in the $SL(2,{\bf Z})$ structure previously seen in the context of vector fields in $AdS_4$, some of the features arising from this exercise, such as the vacuum susceptibility, turns out to be scheme dependent. We explain how scheme independent physical content can be extracted in spite of some degree of scheme dependence in certain quantities.
• ### Passive directors in turbulence(1707.06037)

Feb. 12, 2019 physics.flu-dyn
In experiments and numerical simulations we measured angles between the symmetry axes of small spheroids advected in turbulence ("passive directors"). Since turbulent strains tend to align nearby spheroids, one might think that their relative angles are quite small. We show that this intuition fails in general because angles between the symmetry axes of nearby particles are anomalously large. We identify two mechanisms that cause this phenomenon. First, the dynamics evolves to a fractal attractor despite the fact that the fluid velocity is spatially smooth at small scales. Second, this fractal forms steps akin to scar lines observed in the director patterns for random or chaotic two-dimensional maps.
• ### Crossover from itinerant to localized magnetic excitations through the metal-insulator transition in NaOsO$_{\text{3}}$(1707.05551)

Feb. 4, 2019 cond-mat.str-el
NaOsO$_{\text{3}}$ undergoes a metal-insulator transition (MIT) at 410 K, concomitant with the onset of antiferromagnetic order. The excitation spectra have been investigated through the MIT by resonant inelastic x-ray scattering (RIXS) at the Os L$_{\text{3}}$ edge. Low resolution ($\Delta E \sim$ 300 meV) measurements over a wide range of energies reveal that local electronic excitations do not change appreciably through the MIT. This is consistent with a picture in which structural distortions do not drive the MIT. In contrast, high resolution ($\Delta E \sim$ 56 meV) measurements show that the well-defined, low energy magnons in the insulating state weaken and dampen upon approaching the metallic state. Concomitantly, a broad continuum of excitations develops which is well described by the magnetic fluctuations of a nearly antiferromagnetic Fermi liquid. By revealing the continuous evolution of the magnetic quasiparticle spectrum as it changes its character from itinerant to localized, our results provide unprecedented insight into the nature of the MIT in \naoso. In particular, the presence of weak correlations in the paramagnetic phase implies a degree of departure from the ideal Slater limit.
• ### The Abelian Sandpile Model on Fractal Graphs(1602.03424)

Jan. 17, 2019 math.CO, math.PR
We study the Abelian sandpile model (ASM), a process where grains of sand are placed on a graph's vertices. When the number of grains on a vertex is at least its degree, one grain is distributed to each neighboring vertex. This model has been shown to form fractal patterns on the integer lattice, and using these fractal patterns as motivation, we consider the model on graph approximations of post critically finite (p.c.f) fractals. We determine asymptotic behavior of the diameter of sites toppled and characterize graphs which exhibit a periodic number of grains with respect to the initial placement.
• ### New results on sum-product type growth over fields(1702.01003)

May 19, 2019 math.CO, math.NT
We prove a range of new sum-product type growth estimates over a general field $\mathbb{F}$, in particular the special case $\mathbb{F}=\mathbb{F}_p$. They are unified by the theme of "breaking the $3/2$ threshold", epitomising the previous state of the art. These estimates stem from specially suited applications of incidence bounds over $\mathbb{F}$, which apply to higher moments of representation functions. We establish the estimate $|R[A]| \gtrsim |A|^{8/5}$ for cardinality of the set $R[A]$ of distinct cross-ratios defined by triples of elements of a (sufficiently small if $\mathbb{F}$ has positive characteristic, similarly for the rest of the estimates) set $A\subset \mathbb{F}$, pinned at infinity. The cross-ratio naturally arises in various sum-product type questions of projective nature and is the unifying concept underlying most of our results. It enables one to take advantage of its symmetry properties as an onset of growth of, for instance, products of difference sets. The geometric nature of the cross-ratio enables us to break the version of the above threshold for the minimum number of distinct triangle areas $Ouu'$, defined by points $u,u'$ of a non-collinear point set $P\subset \mathbb{F}^2$. Another instance of breaking the threshold is showing that if $A$ is sufficiently small and has additive doubling constant $M$, then $|AA|\gtrsim M^{-2}|A|^{14/9}$. This result has a second moment version, which allows for new upper bounds for the number of collinear point triples in the set $A\times A\subset \mathbb{F}^2$, the quantity often arising in applications of geometric incidence estimates.
• ### Testing hypotheses via a mixture estimation model(1412.2044)

Dec. 31, 2018 stat.ME
We consider a novel paradigm for Bayesian testing of hypotheses and Bayesian model comparison. Our alternative to the traditional construction of posterior probabilities that a given hypothesis is true or that the data originates from a specific model is to consider the models under comparison as components of a mixture model. We therefore replace the original testing problem with an estimation one that focus on the probability weight of a given model within a mixture model. We analyze the sensitivity on the resulting posterior distribution on the weights of various prior modeling on the weights. We stress that a major appeal in using this novel perspective is that generic improper priors are acceptable, while not putting convergence in jeopardy. Among other features, this allows for a resolution of the Lindley-Jeffreys paradox. When using a reference Beta B(a,a) prior on the mixture weights, we note that the sensitivity of the posterior estimations of the weights to the choice of a vanishes with the sample size increasing and avocate the default choice a=0.5, derived from Rousseau and Mengersen (2011). Another feature of this easily implemented alternative to the classical Bayesian solution is that the speeds of convergence of the posterior mean of the weight and of the corresponding posterior probability are quite similar.
• ### Multitask learning and benchmarking with clinical time series data(1703.07771)

Aug. 9, 2019 cs.LG, stat.ML
Health care is one of the most exciting frontiers in data mining and machine learning. Successful adoption of electronic health records (EHRs) created an explosion in digital clinical data available for analysis, but progress in machine learning for healthcare research has been difficult to measure because of the absence of publicly available benchmark data sets. To address this problem, we propose four clinical prediction benchmarks using data derived from the publicly available Medical Information Mart for Intensive Care (MIMIC-III) database. These tasks cover a range of clinical problems including modeling risk of mortality, forecasting length of stay, detecting physiologic decline, and phenotype classification. We propose strong linear and neural baselines for all four tasks and evaluate the effect of deep supervision, multitask training and data-specific architectural modifications on the performance of neural models.
• Aims: To investigate the extension of the very-high-energy spectral tail of the Crab pulsar at energies above 400 GeV. Methods: We analyzed $\sim$320 hours of good quality data of Crab with the MAGIC telescope, obtained from February 2007 until April 2014. Results: We report the most energetic pulsed emission ever detected from the Crab pulsar reaching up to 1.5 TeV. The pulse profile shows two narrow peaks synchronized with the ones measured in the GeV energy range. The spectra of the two peaks follow two different power-law functions from 70 GeV up to 1.5 TeV and connect smoothly with the spectra measured above 10 GeV by the Large Area Telescope (LAT) on board of the Fermi satellite. When making a joint fit of the LAT and MAGIC data, above 10 GeV, the photon indices of the spectra differ by 0.5$\pm$0.1. Conclusions: We measured with the MAGIC telescopes the most energetic pulsed photons from a pulsar to date. Such TeV pulsed photons require a parent population of electrons with a Lorentz factor of at least $5\times 10^6$. These results strongly suggest IC scattering off low energy photons as the emission mechanism and a gamma-ray production region in the vicinity of the light cylinder.
• ### Higher Order Reentrant Post Modes in Cylindrical Cavities(1611.08939)

Dec. 3, 2018 physics.ins-det
Reentrant cavities are microwave resonant devices employed in a number of different areas of physics. They are appealing due to their simple frequency tuning mechanism, which offers large tuning ranges. Reentrant cavities are, in essence, 3D lumped LC circuits consisting of a conducting central post embedded in a resonant cavity. The lowest order reentrant mode (which transforms from the $TM_{010}$ mode) has been extensively studied in past publications. In this work we show the existence of higher order reentrant post modes (which transform from the $TM_{01n}$ mode family). We characterize these new modes in terms of their frequency tuning, filling factors and quality factors, as well as discuss some possible applications of these modes in fundamental physics tests. The appendix contains a comment on a paper related to this work.
• ### NASCUP: Nucleic Acid Sequence Classification by Universal Probability(1511.04944)

Nov. 29, 2018 cs.IT, math.IT, q-bio.GN
Motivated by the need for fast and accurate classification of unlabeled nucleotide sequences on a large scale, we developed NASCUP, a new classification method that captures statistical structures of nucleotide sequences by compact context-tree models and universal probability from information theory. NASCUP achieved BLAST-like classification accuracy consistently for several large-scale databases in orders-of-magnitude reduced runtime, and was applied to other bioinformatics tasks such as outlier detection and synthetic sequence generation.
• ### ML and Near-ML Decoding of LDPC Codes Over the BEC: Bounds and Decoding Algorithms(1709.01455)

Nov. 20, 2018 cs.IT, math.IT
The performance of maximum-likelihood (ML) decoding on the binary erasure channel for finite-length low-density parity-check (LDPC) codes from two random ensembles is studied. The theoretical average spectrum of the Gallager ensemble is computed by using a recurrent procedure and compared to the empirically found average spectrum for the same ensemble as well as to the empirical average spectrum of the Richardson-Urbanke ensemble and spectra of selected codes from both ensembles. Distance properties of the random codes from the Gallager ensemble are discussed. A tightened union-type upper bound on the ML decoding error probability based on the precise coefficients of the average spectrum is presented. A new upper bound on the ML decoding performance of LDPC codes from the Gallager ensemble based on computing the rank of submatrices of the code parity-check matrix is derived. A new low-complexity near-ML decoding algorithm for quasi-cyclic LDPC codes is proposed and simulated. Its performance is compared to the upper bounds on the ML decoding performance.
• ### Reference Pulse Attack on Continuous-Variable Quantum Key Distribution with Local Local Oscillator under trusted phase noise(1709.10202)

Nov. 16, 2018 quant-ph
We show that partially trusting the phase noise associated with estimation uncertainty in a LLO CVQKD system allows one to exchange higher secure key rates than in the case of untrusted phase noise. However, this opens a security loophole through the manipulation of the reference pulse amplitude. We label this as "reference pulse attack" which is applicable to all LLO-CVQKD systems if the phase noise is trusted. We show that, at the optimal reference pulse intensity level, Eve achieves unity attack efficiency at 23.8km and 32.0km while using lossless and 0.14dB/km loss channels, respectively, for her attack. However, in order to maintain the performance enhancement from partially trusting the phase noise, countermeasures have been proposed. As a result, the LLO-CVQKD system with partially trusted phase noise owns a superior key rate at 20km by an order 9.5, and extended transmission distance by 45%, than that of the phase noise untrusted system.
• ### Observation of Nonlocality Sharing among Three Observers with One Entangled Pair via Optimal Weak Measurement(1609.01863)

Nov. 7, 2018 quant-ph
Bell nonlocality plays a fundamental role in quantum theory. Numerous tests of the Bell inequality have been reported since the ground-breaking discovery of the Bell theorem.Up to now, however, most discussions of the Bell scenario have focused on a single pair of entangled particles distributed to only two separated observers. Recently, it has been shown surprisingly that multiple observers can share the nonlocality present in a single particle from an entangled pair using the method of weak measurements [Phys. Rev. Lett. {\bf 114}, 250401 (2015)]. Here we report an observation of double CHSH-Bell inequality violations for a single pair of entangled photons with strength continuous-tunable optimal weak measurements in photonic system for the first time. Our results not only shed new light on the interplay between nonlocality and quantum measurements but may also be significant for important applications such as unbounded randomness certification and quantum steering.
• ### Social media affects the timing, location, and severity of school shootings(1506.06305)

Nov. 5, 2018 physics.soc-ph, cs.SI
Over the past two decades, school shootings within the United States have repeatedly devastated communities and shaken public opinion. Many of these attacks appear to be `lone wolf' ones driven by specific individual motivations, and the identification of precursor signals and hence actionable policy measures would thus seem highly unlikely. Here, we take a system-wide view and investigate the timing of school attacks and the dynamical feedback with social media. We identify a trend divergence in which college attacks have continued to accelerate over the last 25 years while those carried out on K-12 schools have slowed down. We establish the copycat effect in school shootings and uncover a statistical association between social media chatter and the probability of an attack in the following days. While hinting at causality, this relationship may also help mitigate the frequency and intensity of future attacks.
• ### Cosmological scenarios from multiquintessence(1612.08386)

Oct. 29, 2018 hep-th
In this work we derive and analyse cosmological scenarios coming from multi-component scalar field models. We consider a direct sum of a sine-Gordon with a Z2 model, and also a combination of those with a BNRT model. Moreover, we work with a modified version of the BNRT model, which breaks the Z2 x Z2 symmetry of the original BNRT potential, coupled with the sine-Gordon and with the standard Z2 models. We show that our approach can be straightforwardly elevated to $N$ fields. All the computations are made analytically and some parameters restriction is put forward in order to get in touch with complete and realistic cosmological scenarios.
• ### Computational Topology Techniques for Characterizing Time-Series Data(1708.09359)

Oct. 12, 2018 cs.CG
Topological data analysis (TDA), while abstract, allows a characterization of time-series data obtained from nonlinear and complex dynamical systems. Though it is surprising that such an abstract measure of structure - counting pieces and holes - could be useful for real-world data, TDA lets us compare different systems, and even do membership testing or change-point detection. However, TDA is computationally expensive and involves a number of free parameters. This complexity can be obviated by coarse-graining, using a construct called the witness complex. The parametric dependence gives rise to the concept of persistent homology: how shape changes with scale. Its results allow us to distinguish time-series data from different systems - e.g., the same note played on different musical instruments.
• ### Convolutional neural networks automate detection for tracking of submicron scale particles in 2D and 3D(1704.03009)

Oct. 6, 2018 q-bio.QM
Particle tracking is a powerful biophysical tool that requires conversion of large video files into position time series, i.e. traces of the species of interest for data analysis. Current tracking methods, based on a limited set of input parameters to identify bright objects, are ill-equipped to handle the spectrum of spatiotemporal heterogeneity and poor signal-to-noise ratios typically presented by submicron species in complex biological environments. Extensive user involvement is frequently necessary to optimize and execute tracking methods, which is not only inefficient but introduces user bias. To develop a fully automated tracking method, we developed a convolutional neural network for particle localization from image data, comprised of over 6,000 parameters, and employed machine learning techniques to train the network on a diverse portfolio of video conditions. The neural network tracker provides unprecedented automation and accuracy, with exceptionally low false positive and false negative rates on both 2D and 3D simulated videos and 2D experimental videos of difficult-to-track species.
• ### Improved Lower Bounds on Mutual Information Accounting for Nonlinear Signal-Noise Interaction(1606.09176)

Sept. 28, 2018 cs.IT, math.IT, physics.optics
In fiber-optic communications, evaluation of mutual information (MI) is still an open issue due to the unavailability of an exact and mathematically tractable channel model. Traditionally, lower bounds on MI are computed by approximating the (original) channel with an auxiliary forward channel. In this paper, lower bounds are computed using an auxiliary backward channel, which has not been previously considered in the context of fiber-optic communications. Distributions obtained through two variations of the stochastic digital backpropagation (SDBP) algorithm are used as auxiliary backward channels and these bounds are compared with bounds obtained through the conventional digital backpropagation (DBP). Through simulations, higher information rates were achieved with SDBP, {which can be explained by the ability of SDBP to account for nonlinear signal--noise interactions
• ### Extensive deep neural networks for transferring small scale learning to large scale systems(1708.06686)

We present a physically-motivated topology of a deep neural network that can efficiently infer extensive parameters (such as energy, entropy, or number of particles) of arbitrarily large systems, doing so with O(N) scaling. We use a form of domain decomposition for training and inference, where each sub-domain (tile) is comprised of a non-overlapping focus region surrounded by an overlapping context region. The size of these regions is motivated by the physical interaction length scales of the problem. We demonstrate the application of EDNNs to three physical systems: the Ising model and two hexagonal/graphene-like datasets. In the latter, an EDNN was able to make total energy predictions of a 60 atoms system, with comparable accuracy to density functional theory (DFT), in 57 milliseconds. Additionally EDNNs are well suited for massively parallel evaluation, as no communication is necessary during neural network evaluation. We demonstrate that EDNNs can be used to make an energy prediction of a two-dimensional 35.2 million atom system, over 1 square micrometer of material, at an accuracy comparable to DFT, in under 25 minutes. Such a system exists on a length scale visible with optical microscopy and larger than some living organisms.
• ### Generation of high-fidelity quantum control methods for multi-level systems(1708.02634)

Sept. 13, 2018 quant-ph
In recent decades there has been a rapid development of methods to experimentally control individual quantum systems. A broad range of quantum control methods has been developed for two-level systems, however the complexity of multi-level quantum systems make the development of analogous control methods extremely challenging. Here, we exploit the equivalence between multi-level systems with SU(2) symmetry and spin-1/2 systems to develop a technique for generating new robust, high-fidelity, multi-level control methods. As a demonstration of this technique, we develop new adiabatic and composite multi-level quantum control methods and experimentally realise these methods using an $^{171}$Yb$^+$ ion system. We measure the average infidelity of the process in both cases to be around $10^{-4}$, demonstrating that this technique can be used to develop high-fidelity multi-level quantum control methods and can, for example, be applied to a wide range of quantum computing protocols including implementations below the fault-tolerant threshold in trapped ions.
• ### Quantifiable simulation of quantum computation beyond stochastic ensemble computation(1604.07517)

Aug. 24, 2018 quant-ph
In this study, a distinctive feature of quantum computation (QC) is characterized. To this end, a seemingly-powerful classical computing model, called "stochastic ensemble machine (SEnM)," is considered. The SEnM runs with an ensemble consisting of finite copies of a single probabilistic machine, hence is as powerful as a probabilistic Turing machine (PTM). Then the hypothesis--that is, the SEnM can effectively simulate a general circuit model of QC--is tested by introducing an information-theoretic inequality, named readout inequality. The inequality is satisfied by the SEnM and imposes a critical condition: if the hypothesis holds, the inequality should be satisfied by the probing model of QC. However, it is shown that the above hypothesis is not generally accepted with the inequality violation, namely, such a simulation necessarily fails, implying that PTM $\subseteq$ QC.
• ### Zooming in to Massive Star Birth(1701.05953)

Aug. 17, 2018 astro-ph.GA, astro-ph.SR
We present high resolution (0.2", 1000 AU) 1.3 mm ALMA observations of massive infrared dark cloud clump, G028.37+00.07-C1, thought to harbor the early stages of massive star formation. Using $\rm N_2D^+$(3-2) we resolve the previously identified C1-S core, separating the bulk of its emission from two nearby protostellar sources. C1-S is thus identified as a massive ($\sim50\:M_\odot$), compact ($\sim0.1\:$pc diameter) starless core, e.g., with no signs of outflow activity. Being highly deuterated, this is a promising candidate for a pre-stellar core on the verge of collapse. An analysis of its dynamical state indicates a sub-virial velocity dispersion compared to a trans-Alfv\'enic turbulent core model. However, virial equilibrium could be achieved with sub-Alfv\'enic conditions involving $\sim2\:$mG magnetic field strengths.
• ### Dynamical phase transitions in sampling complexity(1703.05332)

We make the case for studying the complexity of approximately simulating (sampling) quantum systems for reasons beyond that of quantum computational supremacy, such as diagnosing phase transitions. We consider the sampling complexity as a function of time $t$ due to evolution generated by spatially local quadratic bosonic Hamiltonians. We obtain an upper bound on the scaling of $t$ with the number of bosons $n$ for which approximate sampling is classically efficient. We also obtain a lower bound on the scaling of $t$ with $n$ for which any instance of the boson sampling problem reduces to this problem and hence implies that the problem is hard, assuming the conjectures of Aaronson and Arkhipov [Proc. 43rd Annu. ACM Symp. Theory Comput. STOC '11]. This establishes a dynamical phase transition in sampling complexity. Further, we show that systems in the Anderson-localized phase are always easy to sample from at arbitrarily long times. We view these results in the light of classifying phases of physical systems based on parameters in the Hamiltonian. In doing so, we combine ideas from mathematical physics and computational complexity to gain insight into the behavior of condensed matter, atomic, molecular and optical systems.
• ### Vacuum fluctuations of a scalar field near a reflecting boundary and their effects on the motion of a test particle(1709.10392)

Aug. 2, 2018 quant-ph, hep-th
The contribution from quantum vacuum fluctuations of a real massless scalar field to the motion of a test particle that interacts with the field in the presence of a perfectly reflecting flat boundary is here investigated. There is no quantum induced dispersions on the motion of the particle when it is alone in the empty space. However, when a reflecting wall is introduced, dispersions occur with magnitude dependent on how fast the system evolves between the two scenarios. A possible way of implementing this process would be by means of an idealized sudden switching, for which the transition occurs instantaneously. Although the sudden process is a simple and mathematically convenient idealization it brings some divergences to the results, particularly at a time corresponding to a round trip of a light signal between the particle and the wall. It is shown that the use of smooth switching functions, besides regularizing such divergences, enables us to better understand the behavior of the quantum dispersions induced on the motion of the particle. Furthermore, the action of modifying the vacuum state of the system leads to a change in the particle energy that depends on how fast the transition between these states is implemented. Possible implications of these results to the similar case of an electric charge near a perfectly conducting wall are discussed.