arXiv subjects

Yong Han

Publications and source records attributed to Yong Han.

33 records · Page 2Linked to original sources

arXiv2021

Complete weighted Bergman spaces have bounded point evaluations

Let $Ω\subset \mathbb{C}$ be an arbitrary domain in the one-dimensional complex plane equipped with a positive Radon measure $μ$. For any $1\le p< \infty$, it is shown that the weighted Bergman space $A^p(Ω, μ)$ of holomorphic functions is a Banach space if and only if $A^p(Ω, μ)$ has locally uniformly bounded point evaluations. In particular, in the case $p =2$, any complete Bergman space $A^2(Ω, μ)$ is automatically a reproducing kernel Hilbert space.

math.FA↗

arXiv2021

Explaining neural network predictions of material strength

We recently developed a deep learning method that can determine the critical peak stress of a material by looking at scanning electron microscope (SEM) images of the material's crystals. However, it has been somewhat unclear what kind of image features the network is keying off of when it makes its prediction. It is common in computer vision to employ an explainable AI saliency map to tell one what parts of an image are important to the network's decision. One can usually deduce the important features by looking at these salient locations. However, SEM images of crystals are more abstract to the human observer than natural image photographs. As a result, it is not easy to tell what features are important at the locations which are most salient. To solve this, we developed a method that helps us map features from important locations in SEM images to non-abstract textures that are easier to interpret.

eess.IV↗

arXiv2021

Thermodynamics and kinetics of H adsorption and intercalation for graphene on 6H-SiC(0001) from first-principles calculations

Previous experimental observations for H intercalation under graphene on SiC surfaces motivate clarification of configuration stabilities and kinetic processes related to intercalation. From first-principles density-functional-theory (DFT) calculations, we analyze H adsorption and intercalation for graphene on a 6H-SiC(0001) surface, where the system includes two single-atom-thick graphene layers: the top-layer graphene (TLG) and the underling buffer-layer graphene (BLG) above the terminal Si layer. Our chemical potential analysis shows that, in the low-H coverage regime (described by a single H atom within a sufficiently large supercell), intercalation into the gallery between TLG and BLG, or into the gallery underneath BLG, is more favorable thermodynamically than adsorption on top of TLG. However, intercalation into the gallery between TLG and BLG is most favorable. We obtain energy barriers of about 1.3 eV and 2.3 eV for a H atom diffusing on and under TLG, respectively. From an additional analysis of the energy landscape in the vicinity of a step on the TLG, we assess how readily one guest H atom on the TLG terrace can directly penetrate the TLG into the gallery between TLG and BLG versus crossing a TLG step to access the gallery. We also perform DFT calculations for higher H coverages revealing a shift in favorability to intercalation of H underneath BLG, as well as characterizing the variation with H coverage in interlayer spacings.

cond-mat.mes-hall↗

arXiv2021

Boundedness of Gaussian random sums on trees

Let $\mathcal{T}$ be a rooted tree endowed with the natural partial order $\preceq$. Let $(Z(v))_{v\in \mathcal{T}}$ be a sequence of independent standard Gaussian random variables and let $α= (α_k)_{k=1}^\infty$ be a sequence of real numbers with $\sum_{k=1}^\infty α_k^2<\infty$. Set $α_0 =0$ and define a Gaussian process on $\mathcal{T}$ in the following way: \[ G(\mathcal{T}, α; v): = \sum_{u\preceq v} α_{|u|} Z(u), \quad v \in \mathcal{T}, \] where $|u|$ denotes the graph distance between the vertex $u$ and the root vertex. Under mild assumptions on $\mathcal{T}$, we obtain a necessary and sufficient condition for the almost sure boundedness of the above Gaussian process. Our condition is also necessary and sufficient for the almost sure uniform convergence of the Gaussian process $G(\mathcal{T}, α; v)$ along all rooted geodesic rays in $\mathcal{T}$.

math.PR↗

arXiv2021

Reliable Graph Neural Network Explanations Through Adversarial Training

Graph neural network (GNN) explanations have largely been facilitated through post-hoc introspection. While this has been deemed successful, many post-hoc explanation methods have been shown to fail in capturing a model's learned representation. Due to this problem, it is worthwhile to consider how one might train a model so that it is more amenable to post-hoc analysis. Given the success of adversarial training in the computer vision domain to train models with more reliable representations, we propose a similar training paradigm for GNNs and analyze the respective impact on a model's explanations. In instances without ground truth labels, we also determine how well an explanation method is utilizing a model's learned representation through a new metric and demonstrate adversarial training can help better extract domain-relevant insights in chemistry.

cs.LG↗

arXiv2021

A note on the GFF with one free boundary condition

In this note, we shall prove an explicit formula on the probability of the level line of the Gaussian Free Field (GFF) with mixed boundary condition terminating at the free boundary, which generalizes the results of GFF with Dirichlet boundary condition.

math.PR↗

arXiv2020

On a class of random sets of positive integers

In this note, we study a class of random subsets of positive integers induced by Bernoulli random variables. We obtain sufficient conditions such that the random set is almost surely lacunary, does not have bounded gaps and contains infinitely many arithmetic progressions, respectively.

math.PR↗

arXiv2019

Generative Counterfactual Introspection for Explainable Deep Learning

In this work, we propose an introspection technique for deep neural networks that relies on a generative model to instigate salient editing of the input image for model interpretation. Such modification provides the fundamental interventional operation that allows us to obtain answers to counterfactual inquiries, i.e., what meaningful change can be made to the input image in order to alter the prediction. We demonstrate how to reveal interesting properties of the given classifiers by utilizing the proposed introspection approach on both the MNIST and the CelebA dataset.

cs.LG↗

arXiv2018

Defect-Fluorite Gd2Zr2O7 Ceramics under Helium Irradiation: Amorphization, Cell Volume Expansion, and Multi-stage Bubble Formation

Here, we report a study on the radiation resistance enhancement of Gd2Zr2O7 nanograin ceramics, in which amorphization, cell volume expansion and multi-stage helium (He) bubble formation are investigated and discussed. Gd2Zr2O7 ceramics with a series of grain sizes (55-221 nm) were synthesized and irradiated by 190 keV He ion beam up to a fluence of 5x10^17 ions/cm2. Both the degree of post irradiation cell volume expansion and the amorphization fraction appear to be size dependent. As the average grain size evolves from 55 to 221 nm, the degree of post irradiation cell volume expansion increases from 0.56 to 1.02 %, and the amorphization fraction increases from 6.8 to 11.1 %. Additionally, the threshold He concentrations (at. %) of bubbles at different formation stages and locations, including (1) bubbles at grain boundary, (2) bubble-chains and (3) ribbon-like bubbles within the grain, are all found to be much higher in the nanograin ceramic (55 nm) compared with that of the submicron sample (221 nm). We conclude that grain boundary plays a critical role in minimizing the structural defects, and inhibiting the multi-stage He bubble formation process.

cond-mat.mtrl-sci↗

arXiv2017

On The Brownian Loop Measure

In 2003 Lawler and Werner introduced the Brownian loop measure and studied some of its properties. Cardy and Gamsa has predicted a formula for the total mass of the Brownian loop measure on the set of simple loops in the upper half plane and disconnect two given points from the boundary. In this paper we give a rigorous proof of the formula using a result by Beliaev and Viklund and heavy computations.

math-ph↗

arXiv2016

Tailoring Kinetics on a Topological Insulator Surface by Defect-Induced Strain: Pb Mobility on Bi2Te3

Heteroepitaxial structures based on Bi$_{2}$Te$_{3}$-type topological insulators (TIs) exhibit exotic quantum phenomena. For optimal characterization of these phenomena, it is desirable to control the interface structure during film growth on such TIs. In this process, adatom mobility is a key factor. We demonstrate that Pb mobility on the Bi$_{2}$Te$_{3}$(111) surface can be modified by the engineering local strain, ε, which is induced around the point-like defects intrinsically forming in the Bi$_{2}$Te$_{3}$(111) thin film grown on a Si(111)-7 $\times$ 7 substrate. Scanning tunneling microscopy observations of Pb adatom and cluster distributions and first-principles density functional theory (DFT) analyses of the adsorption energy and diffusion barrier E$_{d}$ of Pb adatom on Bi$_{2}$Te$_{3}$(111) surface show a significant influence of ε. Surprisingly, E$_d$ reveals a cusp-like dependence on ε due to a bifurcation in the position of the stable adsorption site at the critical tensile strain ε$_{c}$ $ \approx $ 0.8%. This constitutes a very different strain-dependence of diffusivity from all previous studies focusing on conventional metal or semiconductor surfaces. Kinetic Monte Carlo simulations of Pb deposition, diffusion, and irreversible aggregation incorporating the DFT results reveal adatom and cluster distributions compatible with our experimental observations.

cond-mat.mtrl-sci↗

arXiv2015

Preprint ARPPS Augmented Reality Pipeline Prospect System

This is the preprint version of our paper on ICONIP. Outdoor augmented reality geographic information system (ARGIS) is the hot application of augmented reality over recent years. This paper concludes the key solutions of ARGIS, designs the mobile augmented reality pipeline prospect system (ARPPS), and respectively realizes the machine vision based pipeline prospect system (MVBPPS) and the sensor based pipeline prospect system (SBPPS). With the MVBPPS's realization, this paper studies the neural network based 3D features matching method.

cs.CV↗

arXiv2014

A new method to suppress high-order harmonics for synchrotron radiation soft x-ray beamline

A feasible and convenient method has been proposed to suppress higher-harmonics for varied-line-spacing (VLS) plane grating monochromator in soft x-ray region. Related calculations and experiments demonstrate that decreasing the included angle slightly by changing the parameter of exit arm length can significantly improve light purity. This method is suitable and has been used for experiments of detector calibration in beamline 4B7B at Beijing Synchrotron Radiation Facility (BSRF).

physics.optics↗

arXiv2014

Analysis of magic lengths in growth of supported metallic nanowires

Metallic nanowires can exhibit fascinating physical properties. These unique properties often originate primarily from the quantum confinement of free electrons in a potential well, while electron-electron interactions do not play a decisive role. A recent experimental study shows that self-assembled Ir nanowires grown on Ge(001) surface have a strong length preference: the nanowire lengths are an integer multiple of 4.8 nm. In this paper, a free electron-gas model for geometries corresponding to the nanowires is used to analyze the selection of these preferred or magic lengths. The model shows that the inclusion of even numbers of free electrons in an Ir nanowire corresponds to these magic lengths once an electron spillage effect is taken into account. The model also shows that the stability of the nanowire diminishes with its increasing length, and consequently suggests why no long nanowires are observed in experiments. It is also shown that applying generic results for quantum size effects in a nanofilm geometry is not adequate to accurately describe the length selection in the rather different nanowire geometry, where the transverse dimensions smaller than the electron Fermi wavelength. Finally, the monatomic Au chain growth on Ge(001) surface is also analyzed. In contrast to Ir nanowires, the model shows that the stability of an Au chain strongly depends on its electron spillage.

cond-mat.mes-hall↗

arXiv2012

Interplay between Quantum Size Effect and Strain Effect on Growth of Nanoscale Metal Thin Film

We develop a theoretical framework to investigate the interplay between quantum size effect (QSE) and strain effect on the stability of metal nanofilms. The QSE and strain effect are shown to be coupled through the concept of "quantum electronic stress. First-principles calculations reveal large quantum oscillations in the surface stress of metal nanofilms as a function of film thickness. This adds extrinsically additional strain-coupled quantum oscillations to surface energy of strained metal nanofilms. Our theory enables a quantitative estimation of the amount of strain in experimental samples, and suggests strain be an important factor contributing to the discrepancies between the existing theories and experiments.

cond-mat.mtrl-sci↗