arXiv subjects

Xinxin Li

Publications and source records attributed to Xinxin Li.

27 records · Page 2Linked to original sources

arXiv2022

A semantic hierarchical graph neural network for text classification

The key to the text classification task is language representation and important information extraction, and there are many related studies. In recent years, the research on graph neural network (GNN) in text classification has gradually emerged and shown its advantages, but the existing models mainly focus on directly inputting words as graph nodes into the GNN models ignoring the different levels of semantic structure information in the samples. To address the issue, we propose a new hierarchical graph neural network (HieGNN) which extracts corresponding information from word-level, sentence-level and document-level respectively. Experimental results on several benchmark datasets achieve better or similar results compared to several baseline methods, which demonstrate that our model is able to obtain more useful information for classification from samples.

cs.CL↗

arXiv2021

A Semismooth Newton-Type Method for the Nearest Doubly Stochastic Matrix Problem

We study a semismooth Newton-type method for the nearest doubly stochastic matrix problem where both differentiability and nonsingularity of the Jacobian can fail. The optimality conditions for this problem are formulated as a system of strongly semismooth functions. We show that the so-called local error bound condition does not hold for this system. Thus the guaranteed convergence rate of Newton-type methods is at most superlinear. By exploiting the problem structure, we construct a modified two step semismooth Newton method that guarantees a nonsingular Jacobian matrix at each iteration, and that converges to the nearest doubly stochastic matrix quadratically. To the best of our knowledge, this is the first Newton-type method which converges $Q$-quadratically in the absence of the local error bound condition.

math.OC↗

arXiv2020

A Restricted Dual Peaceman-Rachford Splitting Method for QAP

We revisit and strengthen splitting methods for solving doubly nonnegative, DNN, relaxations of the quadratic assignment problem, QAP. We use a modified restricted contractive splitting method, PRSM, approach. Our strengthened bounds and new dual multiplier estimates improve on the bounds and convergence results in the literature.

math.OC↗

arXiv2019

Hotel2vec: Learning Attribute-Aware Hotel Embeddings with Self-Supervision

We propose a neural network architecture for learning vector representations of hotels. Unlike previous works, which typically only use user click information for learning item embeddings, we propose a framework that combines several sources of data, including user clicks, hotel attributes (e.g., property type, star rating, average user rating), amenity information (e.g., the hotel has free Wi-Fi or free breakfast), and geographic information. During model training, a joint embedding is learned from all of the above information. We show that including structured attributes about hotels enables us to make better predictions in a downstream task than when we rely exclusively on click data. We train our embedding model on more than 40 million user click sessions from a leading online travel platform and learn embeddings for more than one million hotels. Our final learned embeddings integrate distinct sub-embeddings for user clicks, hotel attributes, and geographic information, providing an interpretable representation that can be used flexibly depending on the application. We show empirically that our model generates high-quality representations that boost the performance of a hotel recommendation system in addition to other applications. An important advantage of the proposed neural model is that it addresses the cold-start problem for hotels with insufficient historical click information by incorporating additional hotel attributes which are available for all hotels.

cs.IR↗

arXiv2019

Efficient Super Resolution For Large-Scale Images Using Attentional GAN

Single Image Super Resolution (SISR) is a well-researched problem with broad commercial relevance. However, most of the SISR literature focuses on small-size images under 500px, whereas business needs can mandate the generation of very high resolution images. At Expedia Group, we were tasked with generating images of at least 2000px for display on the website, four times greater than the sizes typically reported in the literature. This requirement poses a challenge that state-of-the-art models, validated on small images, have not been proven to handle. In this paper, we investigate solutions to the problem of generating high-quality images for large-scale super resolution in a commercial setting. We find that training a generative adversarial network (GAN) with attention from scratch using a large-scale lodging image data set generates images with high PSNR and SSIM scores. We describe a novel attentional SISR model for large-scale images, A-SRGAN, that uses a Flexible Self Attention layer to enable processing of large-scale images. We also describe a distributed algorithm which speeds up training by around a factor of five.

cs.CV↗

arXiv2016

Direct Imaging of Dynamic Glassy Behavior in a Strained Manganite Film

Complex many-body interaction in perovskite manganites gives rise to a strong competition between ferromagnetic metallic and charge ordered phases with nanoscale electronic inhomogeneity and glassy behaviors. Investigating this glassy state requires high resolution imaging techniques with sufficient sensitivity and stability. Here, we present the results of a near-field microwave microscope imaging on the strain driven glassy state in a manganite film. The high contrast between the two electrically distinct phases allows direct visualization of the phase separation. The low temperature microscopic configurations differ upon cooling with different thermal histories. At sufficiently high temperatures, we observe switching between the two phases in either direction. The dynamic switching, however, stops below the glass transition temperature. Compared with the magnetization data, the phase separation was microscopically frozen, while spin relaxation was found in a short period of time.

cond-mat.mes-hall↗

arXiv2013

Batch-fabricated cantilever probes with electrical shielding for nanoscale dielectric and conductivity imaging

This paper presents the design and fabrication of batch-processed cantilever probes with electrical shielding for scanning microwave impedance microscopy. The diameter of the tip apex, which defines the electrical resolution, is less than 50 nm. The width of the stripline and the thicknesses of the insulation dielectrics are optimized for a small series resistance (< 5 W) and a small background capacitance (~ 1 pF), both critical for high sensitivity imaging on various samples. The coaxial shielding ensures that only the probe tip interacts with the sample. The structure of the cantilever is designed to be symmetric to balance the stresses and thermal expansions of different layers so that the cantilever remains straight under variable temperatures. Such shielded cantilever probes produced in the wafer scale will facilitate enormous applications on nanoscale dielectric and conductivity imaging.

physics.ins-det↗

arXiv2012

Unexpected Surface Implanted Layer in Static Random Access Memory Devices Observed by Microwave Impedance Microscope

Real-space mapping of doping concentration in semiconductor devices is of great importance for the microelectronic industry. In this work, a scanning microwave impedance microscope (MIM) is employed to resolve the local conductivity distribution of a static random access memory (SRAM) sample. The MIM electronics can also be adjusted to the scanning capacitance microscopy (SCM) mode, allowing both measurements on the same region. Interestingly, while the conventional SCM images match the nominal device structure, the MIM results display certain unexpected features, which originate from a thin layer of the dopant ions penetrating through the protective layers during the heavy implantation steps.

cond-mat.mes-hall↗

arXiv2001

A More General Quantum Searching Algorithm And the Precise Formula of the Amplitude and the Non-symmetric Effects of Different Rotating Angles

This paper presented two general quantum search algorithms. We derived the iterated formulas and the simpler approximate formulas and the precise formula for the amplitude in the desired state. A mathematical proof of Grover's algorithm being optimal among the algorithms with arbitrary phase rotations was given in this paper. This first reported the non-symmetric effects of different rotating angles, and gave the first-order approximate phase condition when rotating angles are different.

quant-ph↗