arXiv subjects

Yan Sun

Publications and source records attributed to Yan Sun.

At least 379 records · Page 21Linked to original sources

arXiv2015

On Linear Regression for Interval-valued Data in $\mathcal{K}_{\mathcal{C}}\left(\mathbb{R}\right)$

It has been some time since interval-valued linear regression was investigated. In this paper, we focus on linear regression for interval-valued data within the framework of random sets. The model we propose generalizes a series of existing models. We establish important properties of the model in the space of compact convex subsets of $\mathbb{R}$, analogous to those for the classical linear regression. Furthermore, we carry out theoretical investigations into the least squares estimation that is widely used in the literature. A simulation study is presented that supports our theorems. Finally, an application to a climate data set is provided to demonstrate the applicability of our model.

stat.ME↗

arXiv2015

Power Efficient and Secure Full-Duplex Wireless Communication Systems

In this paper, we study resource allocation for a full-duplex (FD) radio base station serving multiple half-duplex (HD) downlink and uplink users simultaneously. The considered resource allocation algorithm design is formulated as a non-convex optimization problem taking into account minimum required receive signal-to-interference-plus-noise ratios (SINRs) for downlink and uplink communication and maximum tolerable SINRs at potential eavesdroppers. The proposed optimization framework enables secure downlink and uplink communication via artificial noise generation in the downlink for interfering the potential eavesdroppers. We minimize the weighted sum of the total downlink and uplink transmit power by jointly optimizing the downlink beamformer, the artificial noise covariance matrix, and the uplink transmit power. We adopt a semidefinite programming (SDP) relaxation approach to obtain a tractable solution for the considered problem. The tightness of the SDP relaxation is revealed by examining a sufficient condition for the global optimality of the solution. Simulation results demonstrate the excellent performance achieved by the proposed scheme and the significant transmit power savings enabled optimization of the artificial noise covariance matrix.

cs.IT↗

arXiv2015

Graphene-like Dirac states and Quantum Spin Hall Insulators in the square-octagonal MX2 (M=Mo, W; X=S, Se, Te) Isomers

We studied the square-octagonal lattice of the transition metal dichalcogenide MX$_2$ (with $M$=Mo, W; $X$=S, Se and Te), as an isomer of the normal hexagonal compound of MX$_2$. By band structure calculations, we observe the graphene-like Dirac band structure in a rectangular lattice of MX$_2$ with nonsymmorphic space group symmetry. Two bands with van Hove singularity points cross each at the Fermi energy, leading to two Dirac cones that locates at opposite momenta. Spin-orbit coupling can open a nontrivial gap at these Dirac points and induce the quantum spin Hall (QSH) phase, the 2D topological insulator. Here, square-octagonal MX$_2$ structures realize the interesting graphene physics, such as Dirac bands and QSH effect, in the transition metal dichalcogenides.

cond-mat.mtrl-sci↗

arXiv2015

Two different topological insulator phases in NaBaBi

By breaking the inversion symmetry of the three-dimensional Dirac metal Na$_3$Bi, we realize topological insulator (TI) phases in the known compound NaBaBi using $ab~initio$ calculations. Two distinct TI phases emerge: one phase is due to band inversion between the Bi $p$ and Na $s$ bands, and the other phase (under pressure) is induced by inversion of the Bi $p$ and Ba $d$ bands. Both phases exhibit Dirac-cone-type surface states, but they have opposite spin textures. In the upper cone, a left-handed spin texture exists for the $s$-$p$ inverted phase (similar to a common TI, e.g., Bi$_2$Se$_3$), whereas a right-handed spin texture appears for the $p$-$d$ inverted phase. NaBaBi presents a prototype model of a TI that exhibits different spin textures in the same material. In addition, NaBaBi may exhibit the introduction of correlation effects to the topological state owing to the existence of $d$ states in the band inversion.

cond-mat.mtrl-sci↗

arXiv2015

Conditional Heteroskedasticity of Return Range Processes

Price range contains important information about the asset volatility, and has long been considered an important indicator for it. In this paper, we propose to jointly model the [low, high] price range as a random interval and introduce an interval-valued GARCH (Int-GARCH) model for the corresponding [low, high] return range process. Model properties are presented under the general framework of random sets, and the parameters are estimated by a metric-based conditional least squares (CLS) method. Our empirical analysis of the daily return range data of Dow Jones component stocks yields very interesting results.

stat.ME↗

arXiv2014

Fast Learning of Relational Dependency Networks

A Relational Dependency Network (RDN) is a directed graphical model widely used for multi-relational data. These networks allow cyclic dependencies, necessary to represent relational autocorrelations. We describe an approach for learning both the RDN's structure and its parameters, given an input relational database: First learn a Bayesian network (BN), then transform the Bayesian network to an RDN. Thus fast Bayes net learning can provide fast RDN learning. The BN-to-RDN transform comprises a simple, local adjustment of the Bayes net structure and a closed-form transform of the Bayes net parameters. This method can learn an RDN for a dataset with a million tuples in minutes. We empirically compare our approach to state-of-the art RDN learning methods that use functional gradient boosting, on five benchmark datasets. Learning RDNs via BNs scales much better to large datasets than learning RDNs with boosting, and provides competitive accuracy in predictions.

cs.LG↗

arXiv2014

A Possible Extension of the Scutum-Centaurus Arm into the Outer Second Quadrant

Combining HI data from the Canadian Galactic Plane Survey~(CGPS) and CO data from the Milky Way Imaging Scroll Painting (MWISP) project, we have identified a new segment of a spiral arm between Galactocentric radii of 15 and 19~kpc that apparently lies beyond the Outer Arm in the second Galactic quadrant. Over most of its length, the arm is 400-600 pc thick in $z$. The new arm appears to be the extension of the distant arm recently discovered by Dame & Thaddeus (2011) as well as the Scutum-Centaurus Arm into the outer second quadrant. Our current survey identified a total of 72 molecular clouds with masses on the order of 10$^2$-10$^4$M$_{\sun}$ that probably lie in the new arm. When all of the available data from the CO molecular clouds are fit, the best$-$fitting spiral model gives a pitch angle of 9.3$^{\circ}$ $\pm$0.7$^{\circ}$.

astro-ph.GA↗

arXiv2014

Computing Multi-Relational Sufficient Statistics for Large Databases

Databases contain information about which relationships do and do not hold among entities. To make this information accessible for statistical analysis requires computing sufficient statistics that combine information from different database tables. Such statistics may involve any number of {\em positive and negative} relationships. With a naive enumeration approach, computing sufficient statistics for negative relationships is feasible only for small databases. We solve this problem with a new dynamic programming algorithm that performs a virtual join, where the requisite counts are computed without materializing join tables. Contingency table algebra is a new extension of relational algebra, that facilitates the efficient implementation of this Möbius virtual join operation. The Möbius Join scales to large datasets (over 1M tuples) with complex schemas. Empirical evaluation with seven benchmark datasets showed that information about the presence and absence of links can be exploited in feature selection, association rule mining, and Bayesian network learning.

cs.LG↗

arXiv2014

Topological Metal of NaBi with Ultralow Lattice Thermal Conductivity and Electron-phonon Superconductivity

By means of first-principles and \emph{ab initio} tight-binding calculations, we found that the compound of NaBi is a three-dimensional non-trivial topological metal. Its topological feature can be confirmed by the presence of band inversion, the derived effective Z$_2$ invariant and the non-trivial surface states with the presence of Dirac cones. Interestingly, our calculations further demonstrated that NaBi exhibits the uniquely combined properties between the electron-phonon coupling superconductivity in nice agreement with recent experimental measurements and the obviously anisotropic but extremely low thermal conductivity. The spin-orbit coupling effects greatly affect those properties. NaBi may provide a rich platform to study the relationship among metal, topology, superconductivity and thermal conductivity.

cond-mat.mtrl-sci↗

arXiv2014

A semiparametric spatial dynamic model

Stimulated by the Boston house price data, in this paper, we propose a semiparametric spatial dynamic model, which extends the ordinary spatial autoregressive models to accommodate the effects of some covariates associated with the house price. A profile likelihood based estimation procedure is proposed. The asymptotic normality of the proposed estimators are derived. We also investigate how to identify the parametric/nonparametric components in the proposed semiparametric model. We show how many unknown parameters an unknown bivariate function amounts to, and propose an AIC/BIC of nonparametric version for model selection. Simulation studies are conducted to examine the performance of the proposed methods. The simulation results show our methods work very well. We finally apply the proposed methods to analyze the Boston house price data, which leads to some interesting findings.

math.ST↗

arXiv2014

A Normal Hierarchical Model and Minimum Contrast Estimation for Random Intervals

Many statistical data are imprecise due to factors such as measurement errors, computation errors, and lack of information. In such cases, data are better represented by intervals rather than by single numbers. Existing methods for analyzing interval-valued data include regressions in the metric space of intervals and symbolic data analysis, the latter being proposed in a more general setting. However, there has been a lack of literature on the distribution-based inferences for interval-valued data. In an attempt to fill this gap, we extend the concept of normality for random sets by Lyashenko (1983) and propose a normal hierarchical model for random intervals. In addition, we develop a minimum contrast estimator (MCE) for the model parameters, which we show is both consistent and asymptotically normal. Simulation studies support our theoretical findings, and show very promising results. Finally, we successfully apply our model and MCE to a real dataset.

math.ST↗

arXiv2013

Simultaneous sparse model selection and coefficient estimation for heavy-tailed autoregressive processes

We propose a sparse coefficient estimation and automated model selection procedure for autoregressive (AR) processes with heavy-tailed innovations based on penalized conditional maximum likelihood. Under mild moment conditions on the innovation processes, the penalized conditional maximum likelihood estimator (PCMLE) satisfies a strong consistency, $O_P(N^{-1/2})$ consistency, and the oracle properties, where N is the sample size. We have the freedom in choosing penalty functions based on the weak conditions on them. Two penalty functions, least absolute shrinkage and selection operator (LASSO) and smoothly clipped average deviation (SCAD), are compared. The proposed method provides a distribution-based penalized inference to AR models, which is especially useful when the other estimation methods fail or under perform for AR processes with heavy-tailed innovations (see \cite{Resnick}). A simulation study confirms our theoretical results. At the end, we apply our method to a historical price data of the US Industrial Production Index for consumer goods, and obtain very promising results.

stat.ME↗

arXiv2013

Rock-salt SnS and SnSe: Native Topological Crystalline Insulators

Unlike time-reversal topological insulators, surface metallic states with Dirac cone dispersion in the recently discovered topological crystalline insulators (TCIs) are protected by crystal symmetry. To date, TCI behaviors have been observed in SnTe and the related alloys Pb$_{1-x}$Sn$_{x}$Se/Te, which incorporate heavy elements with large spin-orbit coupling (SOC). Here, by combining first-principles and {\it ab initio} tight-binding calculations, we report the formation of a TCI in the relatively lighter rock-salt SnS and SnSe. This TCI is characterized by an even number of Dirac cones at the high-symmetry (001), (110) and (111) surfaces, which are protected by the reflection symmetry with respect to the ($\bar{1}$10) mirror plane. We find that both SnS and SnSe have an intrinsically inverted band structure and the SOC is necessary only to open the bulk band gap. The bulk band gap evolution upon volume expansion reveals a topological transition from an ambient pressure TCI to a topologically trivial insulator. Our results indicate that the SOC alone is not sufficient to drive the topological transition.

cond-mat.mtrl-sci↗

arXiv2013

Tuning the vertical location of helical surface states in topological insulator heterostructures via dual-proximity effects

In integrating topological insulators (TIs) with conventional materials, one crucial issue is how the topological surface states (TSS) will behave in such heterostructures. We use first-principles approaches to establish accurate tunability of the vertical location of the TSS via intriguing dual-proximity effects. By depositing a conventional insulator (CI) overlayer onto a TI substrate (Bi2Se3 or Bi2Te3), we demonstrate that, the TSS can float to the top of the CI film, or stay put at the CI/TI interface, or be pushed down deeper into the otherwise structurally homogeneous TI substrate. These contrasting behaviors imply a rich variety of possible quantum phase transitions in the hybrid systems, dictated by key material-specific properties of the CI. These discoveries lay the foundation for accurate manipulation of the real space properties of TSS in TI heterostructures of diverse technological significance.

cond-mat.mtrl-sci↗

arXiv2012

Dirac semimetal and topological phase transitions in A3Bi (A=Na, K, Rb)

The three-dimensional (3D) Dirac point, where two Weyl points overlap in momentum space, is usually unstable and hard to realize. Here we show, based on the first-principles calculations and effective model analysis, that crystalline $A_3$Bi ($A$=Na, K, Rb) are Dirac semimetals with bulk 3D Dirac points protected by crystal symmetry. They possess non-trivial Fermi arcs on the surfaces, and can be driven into various topologically distinct phases by explicit breaking of symmetries. Giant diamagnetism, linear quantum magnetoresistance, and quantum spin-Hall effect will be expected for such compounds.

cond-mat.mtrl-sci↗

arXiv2012

A 95 GHz Class I Methanol Maser Survey Toward A Sample of GLIMPSE Point Sources Associated with BGPS Clumps

We report a survey with the Purple Mountain Observatory (PMO) 13.7-m radio telescope for class I methanol masers from the 95 GHz (8_0 - 7_1 A^+) transition. The 214 target sources were selected by combining information from both the Spitzer GLIMPSE and 1.1 mm BGPS survey catalogs. The observed sources satisfy both the GLIMPSE mid-IR criteria of [3.6]-[4.5]>1.3, [3.6]-[5.8]>2.5, [3.6]-[8.0]>2.5 and 8.0 um magnitude less than 10, and also have an associated 1.1 mm BGPS source. Class I methanol maser emission was detected in 63 sources, corresponding to a detection rate of 29% for this survey. For the majority of detections (43), this is the first identification of a class I methanol maser associated with these sources. We show that the intensity of the class I methanol maser emission is not closely related to mid-IR intensity or the colors of the GLIMPSE point sources, however, it is closely correlated with properties (mass and beam-averaged column density) of the BGPS sources. Comparison of measures of star formation activity for the BGPS sources with and without class I methanol masers indicate that the sources with class I methanol masers usually have higher column density and larger flux density than those without them. Our results predict that the criteria log(S_{int})<-38.0+1.72log(N_{H_{2}}^{beam}) and log(N_{H_{2}}^{beam})>22.1, which utilizes both the integrated flux density (S_{int}) and beam-averaged column density (N_{H_{2}}^{beam}) of the BGPS sources, are very efficient for selecting sources likely to have an associated class I methanol maser. Our expectation is that searches using these criteria will detect 90% of the predicted number of class I methanol masers from the full BGPS catalog (~ 1000), and do so with a high detection efficiency (~75%).

astro-ph.GA↗

arXiv2011

Strain driven onset of non-trivial topological insulating states in Zintl Sr$_2X$ compounds ($X$=Pb, Sn)

We explore the topological behavior of the binary Zintl phase of the alkaline earth metals based compounds Sr$_2$Pb and Sr$_2$Sn using both standard and hybrid density functional theory. It is found that Sr$_2$Pb lies on the verge of a topological instability which can be suitably tuned through the application of a small uniaxial expansion strain ($>$ 3%). The resulting non-trivial topologically insulating state display well-defined metallic states in the Sr$_{2}$Pb(010) surface, whose evolution is studied as a function of the film thickness.

cond-mat.mtrl-sci↗

arXiv2011

C2H, HC3N and HNC Observations in OMC-2/3

For the first time, the OMC-2/3 region was mapped in C2H(1-0), HC3N(10-9) and HNC(1-0) lines. In general, the emissions from all the three molecular species reveal an extended filamentary structure. The distribution of C2H cores almost follows that of the 1300 $μ$m condensations, which might suggest that C2H is a good tracer to study the core structure of molecular clouds. The core masses traced by HNC are rather flat, ranging from 18.8 to 49.5 $M_{\odot}$, while present a large span for those from C2H, ranging from 6.4 to 36.0 $M_{\odot}$. The line widths of both HNC and C2H look very similar, and both are wider than that of HC3N. The line widths of the three lines are all wider than those from dark clouds, implying that the former is more active than the latter, and has larger turbulence caused by winds and UV radiation from the surrounding massive stars.

astro-ph.SR↗