arXiv subjects

Yi Gao

Publications and source records attributed to Yi Gao.

At least 91 records · Page 5Linked to original sources

arXiv2017

Possible spin excitation structure in monolayer FeSe grown on SrTiO$_{3}$

Based on recent high-resolution angle-resolved photoemission spectroscopy measurement in monolayer FeSe grown on SrTiO$_{3}$, we constructed a tight-binding model and proposed a superconducting (SC) pairing function which can well fit the observed band structure and SC gap anisotropy. Then we investigated the spin excitation spectra in order to determine the possible sign structure of the SC order parameter. We found that a resonance-like spin excitation may occur if the SC order parameter changes sign along the Fermi surfaces. However, this resonance is located at different locations in momentum space compared to other FeSe-based superconductors, suggesting that the Fermi surface shape and pairing symmetry in monolayer FeSe grown on SrTiO$_{3}$ may be different from other FeSe-based superconductors.

cond-mat.supr-con↗

arXiv2017

Sparse Autoencoder for Unsupervised Nucleus Detection and Representation in Histopathology Images

Histopathology images are crucial to the study of complex diseases such as cancer. The histologic characteristics of nuclei play a key role in disease diagnosis, prognosis and analysis. In this work, we propose a sparse Convolutional Autoencoder (CAE) for fully unsupervised, simultaneous nucleus detection and feature extraction in histopathology tissue images. Our CAE detects and encodes nuclei in image patches in tissue images into sparse feature maps that encode both the location and appearance of nuclei. Our CAE is the first unsupervised detection network for computer vision applications. The pretrained nucleus detection and feature extraction modules in our CAE can be fine-tuned for supervised learning in an end-to-end fashion. We evaluate our method on four datasets and reduce the errors of state-of-the-art methods up to 42%. We are able to achieve comparable performance with only 5% of the fully-supervised annotation cost.

cs.CV↗

arXiv2017

Au13(8e): A Secondary Block for Describing a Special Group of Liganded Gold Clusters Containing Icosahedral Au13 Motifs

A grand unified model (GUM) has been proposed recently to understand structure anatomy and evolution of liganded gold clusters. In this work, besidesthe two types of elementary blocks (triangular Au3(2e) and tetrahedral Au4(2e)), we introduce a secondary block, namely, the icosahedral Au13 with 8e valence electrons, noted as Au13(8e). Using this secondary block, structural anatomy and evolution of a special group of liganded gold nanoclusters containing icosahedral Au13 motifs can be conveniently analyzed. In addition, a new ligand-protected cluster Au49(PR3)10(SR)15Cl2 is predicted to exhibit high chemical and thermal stability, suggesting likelihood of its synthesis in the laboratory.

cond-mat.mtrl-sci↗

arXiv2016

Neural Networks with Smooth Adaptive Activation Functions for Regression

In Neural Networks (NN), Adaptive Activation Functions (AAF) have parameters that control the shapes of activation functions. These parameters are trained along with other parameters in the NN. AAFs have improved performance of Neural Networks (NN) in multiple classification tasks. In this paper, we propose and apply AAFs on feedforward NNs for regression tasks. We argue that applying AAFs in the regression (second-to-last) layer of a NN can significantly decrease the bias of the regression NN. However, using existing AAFs may lead to overfitting. To address this problem, we propose a Smooth Adaptive Activation Function (SAAF) with piecewise polynomial form which can approximate any continuous function to arbitrary degree of error. NNs with SAAFs can avoid overfitting by simply regularizing the parameters. In particular, an NN with SAAFs is Lipschitz continuous given a bounded magnitude of the NN parameters. We prove an upper-bound for model complexity in terms of fat-shattering dimension for any Lipschitz continuous regression model. Thus, regularizing the parameters in NNs with SAAFs avoids overfitting. We empirically evaluated NNs with SAAFs and achieved state-of-the-art results on multiple regression datasets.

cs.CV↗

arXiv2016

Hidden sign-changing $s$-wave superconductivity in monolayer FeSe

Combining the recent scanning tunneling microscopy (STM) and angle-resolved photoemission spectroscopy (ARPES) measurements, we construct a tight-binding model suitable for describing the band structure of monolayer FeSe grown on SrTiO$_{3}$. Then we propose a possible pairing function, which can well describe the gap anisotropy observed by ARPES and has a hidden sign-changing characteristic. At last, as a test of this pairing function we further study the nonmagnetic impurity-induced bound states, to be verified by future STM experiments.

cond-mat.supr-con↗

arXiv2016

Simulation of optical response functions in molecular junctions

We discuss theoretical approaches to nonlinear optical spectroscopy of molecular junctions. Optical response functions are derived in the form convenient for implementation of Green function techniques, and their expressions in terms of pseudoparticle nonequilibrium Green functions are proposed. The formulation allows to account for both intra-molecular interactions and hybridization of molecular states due to coupling to contacts. Two-dimensional optical spectroscopy in junctions is considered as an example.

cond-mat.mes-hall↗

arXiv2016

On the widths of Stokes lines in Raman scattering from molecules adsorbed at metal surfaces and in molecular conduction junctions

Within a generic model we analyze the Stokes linewidth in surface enhanced Raman scattering (SERS) from molecules embedded as bridges in molecular junctions. We identify four main contributions to the off-resonant Stokes signal and show that under zero voltage bias (a situation pertaining also to standard SERS experiments) and at low bias junctions only one of these contributions is pronounced. The linewidth of this component is determined by the molecular vibrational relaxation rate, which is dominated by interactions with the essentially bosonic thermal environment when the relevant molecular electronic energy is far from the metal(s) Fermi energy(ies). It increases when the molecular electronic level is close to the metal Fermi level so that an additional vibrational relaxation channel due to electron-hole (eh) excition in the molecule opens. Other contributions to the Raman signal, of considerably broader linewidths, can become important at larger junction bias.

cond-mat.mes-hall↗

arXiv2016

Patch-based Convolutional Neural Network for Whole Slide Tissue Image Classification

Convolutional Neural Networks (CNN) are state-of-the-art models for many image classification tasks. However, to recognize cancer subtypes automatically, training a CNN on gigapixel resolution Whole Slide Tissue Images (WSI) is currently computationally impossible. The differentiation of cancer subtypes is based on cellular-level visual features observed on image patch scale. Therefore, we argue that in this situation, training a patch-level classifier on image patches will perform better than or similar to an image-level classifier. The challenge becomes how to intelligently combine patch-level classification results and model the fact that not all patches will be discriminative. We propose to train a decision fusion model to aggregate patch-level predictions given by patch-level CNNs, which to the best of our knowledge has not been shown before. Furthermore, we formulate a novel Expectation-Maximization (EM) based method that automatically locates discriminative patches robustly by utilizing the spatial relationships of patches. We apply our method to the classification of glioma and non-small-cell lung carcinoma cases into subtypes. The classification accuracy of our method is similar to the inter-observer agreement between pathologists. Although it is impossible to train CNNs on WSIs, we experimentally demonstrate using a comparable non-cancer dataset of smaller images that a patch-based CNN can outperform an image-based CNN.

cs.CV↗

arXiv2016

Optical spectroscopy of molecular junctions: Nonequilibrium Green's functions perspective

We consider optical spectroscopy of molecular junctions from the quantum transport perspective when radiation field is quantized and optical response of the system is simulated as photon flux. Using exact expressions for photon and electronic fluxes derived within the nonequilibrium Green function (NEGF) methodology and utilizing fourth order diagrammatic perturbation theory in molecular coupling to radiation field we perform simulations employing realistic parameters. Results of the simulations are compared to the bare perturbation theory (PT) usually employed in studies on nonlinear optical spectroscopy to classify optical processes. We show that the bare PT violates conservation laws, while flux conserving NEGF formulation mixes optical processes.

cond-mat.mes-hall↗

arXiv2016

Water-Induced Bimetallic Alloy Surface Segregation: A First Principle Study

Bimetallic alloys have drawn extensive attentions in materials science due to their widespread applications in electronics, engineering and catalysis. A very fundamental question of alloy is its surface segregation phenomenon. Many recent observations have shown that reactive gases or supports may have strong effects on alloy segregation. However, segregation in water, the most common solvent and environment, has not received enough attention. In this paper we give the quantitative descriptions on the surface segregation energies of 23 transition-metal impurities in Cu hosts under the conditions of water adsorption by performing density functional theory (DFT) calculations. The general trends in the changes of segregation energies caused by water adsorption are established. Our results show water adsorption could induce strong surface segregation tendencies for early and middle transition metals in Cu alloys. This finding not only prompts us to re-examine the potential effects of water on bimetallic alloy surfaces, but would be also very helpful as a guide for the further theoretical and experimental studies in this field.

cond-mat.mtrl-sci↗

arXiv2015

Superconductivity in doped inversion-symmetric Weyl semimetals

We study theoretically the superconductivity in doped Weyl semimetals with an inversion symmetry based on the Bogoliubov-de Gennes equations. In principle, the two superconducting states, i.e., the zero momentum BCS-like pairing and the finite momentum Fulde-Ferrell-Larkin-Ovchinnikov (FFLO) pairing are competing in this kind of systems. Our self-consistent calculation indicates that the BCS-type state may be the ground state. The competition between these two pairing states is studied in detail through normal state Fermi surface and the finite energy spectral functions. Generally, the Fermi surface topology supports the FFLO pairing while the finite energy band structure favors the BCS-type pairing. We also study the physical properties and address the Majorana Fermions excitation in these two superconducting state respectively.

cond-mat.supr-con↗

arXiv2015

Tunneling interstitial impurity in iron-chalcogenide based superconductors

A pronounced local in-gap zero-energy bound state (ZBS) has been observed by recent scanning tunneling microscopy (STM) experiments on the interstitial Fe impurity (IFI) and its nearest-neighboring (nn) sites in $\mathrm{FeTe_{0.5}Se_{0.5}}$ superconducting (SC) compound. By introducing a new impurity mechanism, the so-called tunneling impurity, and based on the Bogoliubove-de Gennes (BDG) equations we investigated the low-lying energy states of the IFI and the underlying Fe-plane. We found the peak of ZBS does not shift or split in a magnetic field as long as the tunneling parameter between IFI and the Fe-plane is sufficiently small and the Fe-plane is deep in the SC state. Our results are in good agreement with the experiments. We also predicted that modulation of spin density wave (SDW), or charge density wave (CDW) will suppress the intensity of the ZBS.

cond-mat.supr-con↗

arXiv2015

Majorana zero modes in the hopping-modulated one-dimensional $p$-wave superconducting model

We investigate the one-dimensional $p$-wave superconducting model with periodically modulated hopping and show that under time-reversal symmetry, the number of the Majorana zero modes (MZMs) strongly depends on the modulation period. If the modulation period is odd, there can be at most one MZM. However if the period is even, the number of the MZMs can be zero, one and two. In addition, the MZMs will disappear as the chemical potential varies. We derive the condition for the existence of the MZMs and show that the topological properties in this model are dramatically different from the one with periodically modulated potential.

cond-mat.supr-con↗

arXiv2015

Possible superconductivity in Sr$_{2}$IrO$_{4}$ probed by quasiparticle interference

Based on the possible superconducting (SC) pairing symmetries recently proposed, the quasiparticle interference (QPI) patterns in electron- and hole-doped Sr$_{2}$IrO$_{4}$ are theoretically investigated. In the electron-doped case, the QPI spectra can be explained based on a model similar to the octet model of the cuprates while in the hole-doped case, both the Fermi surface topology and the sign of the SC order parameter resemble those of the iron pnictides and there exists a QPI vector resulting from the interpocket scattering between the electron and hole pockets. In both cases, the evolution of the QPI vectors with energy and their behaviors in the nonmagnetic and magnetic impurity scattering cases can well be explained based on the evolution of the constant-energy contours and the sign structure of the SC order parameter. The QPI spectra presented in this paper can be compared with future scanning tunneling microscopy experiments to test whether there are SC phases in electron- and hole-doped Sr$_{2}$IrO$_{4}$ and what the pairing symmetry is.

cond-mat.supr-con↗

arXiv2015

Robust $s_{\pm}$-wave superconductivity against multi-impurity in iron-based superconductors

Effects of disorder on electron-doped iron pnictides are investigated systematically based on self-consistent Bogoliubov-de Gennes equations. Multiply impurities with same scattering potential (SP) are randomly distributed in a square lattice. Probability distribution functions of normalized order parameters for different impurity concentrations $δ_{imp}$, different electron doping concentrations $δ$ are investigated for given SPs. Samples are found to be very robust against weak SP, in which order parameters do not have qualitative change even at very large $δ_{imp}$. While strong SP is able to easily break down the order parameters. For moderate SP, variations of order parameters on and around impurities strongly depend on $δ$, however the distribution functions of normalized order parameters have similar behavior as $δ_{imp}$ increases. Compared with superconducting (SC) order, the magnetic order is more sensitive to multi-impurity effect. The spatial spin density wave pattern has already been destroyed before the system loses its superconductivity. Dependence of SC order on temperature is similar to that of impurity-free case, with the critical temperature being remarkably suppressed for high $δ_{imp}$.

cond-mat.supr-con↗

arXiv2014

Sublinear scaling for time-dependent stochastic density functional theory

A stochastic approach to time-dependent density functional theory (TDDFT) is developed for computing the absorption cross section and the random phase approximation (RPA) correlation energy. The core idea of the approach involves time-propagation of a small set of stochastic orbitals which are first projected on the occupied space and then propagated in time according to the time-dependent Kohn-Sham equations. The evolving electron density is exactly represented when the number of random orbitals is infinite, but even a small number (? 16) of such orbitals is enough to obtain meaningful results for absorption spectrum and the RPA correlation energy per electron. We implement the approach for silicon nanocrystals (NCs) using real-space grids and find that the overall scaling of the algorithm is sublinear with computational time and memory.

physics.chem-ph↗

arXiv2014

Charge-impurity-induced Majorana fermions in topological superconductors

We study numerically Majorana fermions (MFs) induced by a charged impurity in topological superconductors. It is revealed from the relevant Bogoliubov-de Gennes equations that (i) for quasi-one dimensional systems, a pair of MFs are bounded at the two sides of one charge impurity and well separated; and (ii) for a two dimensional square lattice, the charged-impurity-induced MFs are similar to the known pair of vortex-induced MFs, in which one MF is bounded by the impurity while the other appears at the boundary. Moreover, the corresponding local density of states is explored, demonstrating that the presence of MF states may be tested experimentally.

cond-mat.supr-con↗

arXiv2014

Mass transportation of thermally driven nanotube nanomotors with defects

Thermally driven nanotube nanomotors provide linear mass transportation controlled by a temperature gradient. However, the underlying mechanism is still unclear where the mass transportation velocity in experiment is much lower than that resulting from simulations. Considering that defects are common in fabricated nanotubes, we use molecular dynamics simulations to show that the mass transportation would be considerably impeded by the potential barriers or wells induced by the defects, which provides a possible picture to understand the relative low value at microscopic level. The optimal structure and the factors which would affect the performance are discussed. The result indicates considering defects is helpful in designing nanotube nanomotor and other new nanomotor-based devices.

cond-mat.mes-hall↗