Searcharxiv⌕ Search

arXiv subjects

Ying Zhang

Publications and source records attributed to Ying Zhang.

At least 343 records · Page 19Linked to original sources
arXiv2023

The Structure of Flavor Mixing and Reconstruction of the Mass Matrix

The Fermion flavor structure is investigated by bilinear decomposition of the mass matrix after EW symmetry breaking, and the roles of factorized matrices in flavor mixing and mass generation are explored. It is shown that flavor mixing can be addressed as an independent issue. On a new Yukawa basis, the minimal parameterization of flavor mixing is realized containing two relative phases and two free $SO(2)_L$ rotation angles. The validity of the flavor mixing structure is checked in both the lepton and quark sectors. Under the decomposition of flavor mixing, fermion mass matrices are reconstructed under the hierarchy limit. A flat mass matrix with all elements equal to 1 arises naturally from the requirement that homology exists between up-type and down-type fermion mass matrices. Some hints of a flat matrix and flavor breaking are also discussed.

hep-ph↗
arXiv2023

LLMCad: Fast and Scalable On-device Large Language Model Inference

Generative tasks, such as text generation and question answering, hold a crucial position in the realm of mobile applications. Due to their sensitivity to privacy concerns, there is a growing demand for their execution directly on mobile devices. Currently, the execution of these generative tasks heavily depends on Large Language Models (LLMs). Nevertheless, the limited memory capacity of these devices presents a formidable challenge to the scalability of such models. In our research, we introduce LLMCad, an innovative on-device inference engine specifically designed for efficient generative Natural Language Processing (NLP) tasks. The core idea behind LLMCad revolves around model collaboration: a compact LLM, residing in memory, takes charge of generating the most straightforward tokens, while a high-precision LLM steps in to validate these tokens and rectify any identified errors. LLMCad incorporates three novel techniques: (1) Instead of generating candidate tokens in a sequential manner, LLMCad employs the smaller LLM to construct a token tree, encompassing a wider range of plausible token pathways. Subsequently, the larger LLM can efficiently validate all of these pathways simultaneously. (2) It employs a self-adjusting fallback strategy, swiftly initiating the verification process whenever the smaller LLM generates an erroneous token. (3) To ensure a continuous flow of token generation, LLMCad speculatively generates tokens during the verification process by implementing a compute-IO pipeline. Through an extensive series of experiments, LLMCad showcases an impressive token generation speed, achieving rates up to 9.3x faster than existing inference engines.

cs.NI↗
arXiv2023

Control of spectral extreme events in ultrafast fibre lasers by a genetic algorithm

Extreme wave events or rogue waves (RWs) are both statistically rare and of exceptionally large amplitude. They are observed in many complex systems ranging from oceanic and optical environments to financial models and Bose-Einstein condensates. As they appear from nowhere and disappear without a trace, their emergence is unpredictable and non-repetitive, which make them particularly challenging to control. Here, we extend the use of genetic algorithms (GAs), which have been exclusively designed for searching and optimising stationary or repetitive processes in nonlinear optical systems, to the active control of extreme events in a fibre laser cavity. Feeding real-time spectral measurements into a GA controlling the electronics to optimise the cavity parameters, we are able to trigger wave events in the cavity that have the typical statistics of RWs in the frequency domain. This accurate control enables the generation of the optical RWs with a spectral peak intensity 32.8 times higher than the significant intensity threshold. A rationale is proposed and confirmed by numerical simulations of the laser model for the related frequency up- and down-shifting of the optical spectrum that are experimentally observed.

physics.optics↗
arXiv2023

A Robust ADMM-Based Optimization Algorithm For Underwater Acoustic Channel Estimation

Accurate estimation of the Underwater acoustic (UWA) is a key part of underwater communications, especially for coherent systems. The severe multipath effects and large delay spreads make the estimation problem large-scale. The non-stationary, non-Gaussian, and impulsive nature of ocean ambient noise poses further obstacles to the design of estimation algorithms. Under the framework of compressed sensing (CS), this work addresses the issue of robust channel estimation when measurements are contaminated by impulsive noise. A first-order algorithm based on alternating direction method of multipliers (ADMM) is proposed. Numerical simulations of time-varying channel estimation are performed to show its improved performance in highly impulsive noise environments.

eess.SP↗
arXiv2023

Prototypical Cross-domain Knowledge Transfer for Cervical Dysplasia Visual Inspection

Early detection of dysplasia of the cervix is critical for cervical cancer treatment. However, automatic cervical dysplasia diagnosis via visual inspection, which is more appropriate in low-resource settings, remains a challenging problem. Though promising results have been obtained by recent deep learning models, their performance is significantly hindered by the limited scale of the available cervix datasets. Distinct from previous methods that learn from a single dataset, we propose to leverage cross-domain cervical images that were collected in different but related clinical studies to improve the model's performance on the targeted cervix dataset. To robustly learn the transferable information across datasets, we propose a novel prototype-based knowledge filtering method to estimate the transferability of cross-domain samples. We further optimize the shared feature space by aligning the cross-domain image representations simultaneously on domain level with early alignment and class level with supervised contrastive learning, which endows model training and knowledge transfer with stronger robustness. The empirical results on three real-world benchmark cervical image datasets show that our proposed method outperforms the state-of-the-art cervical dysplasia visual inspection by an absolute improvement of 4.7% in top-1 accuracy, 7.0% in precision, 1.4% in recall, 4.6% in F1 score, and 0.05 in ROC-AUC.

cs.CV↗
arXiv2023

VoxelMap++: Mergeable Voxel Mapping Method for Online LiDAR(-inertial) Odometry

This paper presents VoxelMap++: a voxel mapping method with plane merging which can effectively improve the accuracy and efficiency of LiDAR(-inertial) based simultaneous localization and mapping (SLAM). This map is a collection of voxels that contains one plane feature with 3DOF representation and corresponding covariance estimation. Considering total map will contain a large number of coplanar features (kid planes), these kid planes' 3DOF estimation can be regarded as the measurements with covariance of a larger plane (father plane). Thus, we design a plane merging module based on union-find which can save resources and further improve the accuracy of plane fitting. This module can distinguish the kid planes in different voxels and merge these kid planes to estimate the father plane. After merging, the father plane 3DOF representation will be more accurate than the kids plane and the uncertainty will decrease significantly which can further improve the performance of LiDAR(-inertial) odometry. Experiments on challenging environments such as corridors and forests demonstrate the high accuracy and efficiency of our method compared to other state-of-the-art methods (see our attached video). By the way, our implementation VoxelMap++ is open-sourced on GitHub which is applicable for both non-repetitive scanning LiDARs and traditional scanning LiDAR.

cs.RO↗
arXiv2023

Bootstrapping Contrastive Learning Enhanced Music Cold-Start Matching

We study a particular matching task we call Music Cold-Start Matching. In short, given a cold-start song request, we expect to retrieve songs with similar audiences and then fastly push the cold-start song to the audiences of the retrieved songs to warm up it. However, there are hardly any studies done on this task. Therefore, in this paper, we will formalize the problem of Music Cold-Start Matching detailedly and give a scheme. During the offline training, we attempt to learn high-quality song representations based on song content features. But, we find supervision signals typically follow power-law distribution causing skewed representation learning. To address this issue, we propose a novel contrastive learning paradigm named Bootstrapping Contrastive Learning (BCL) to enhance the quality of learned representations by exerting contrastive regularization. During the online serving, to locate the target audiences more accurately, we propose Clustering-based Audience Targeting (CAT) that clusters audience representations to acquire a few cluster centroids and then locate the target audiences by measuring the relevance between the audience representations and the cluster centroids. Extensive experiments on the offline dataset and online system demonstrate the effectiveness and efficiency of our method. Currently, we have deployed it on NetEase Cloud Music, affecting millions of users. Code will be released in the future.

cs.IR↗
arXiv2023

Coexistence of Logarithmic and SdH Quantum Oscillations in Ferromagnetic Cr-doped Tellurium Single Crystals

We report the synthesis of transition-metal-doped ferromagnetic elemental single-crystal semiconductors with quantum oscillations using the physical vapor transport method. The 7.7 atom% Cr-doped Te crystals (Cr_Te) show ferromagnetism, butterfly-like negative magnetoresistance in the low temperature (< 3.8 K) and low field (< 0.15 T) region, and high Hall mobility, e.g., 1320 cm2 V-1 s-1 at 30 K and 350 cm2 V-1 s-1 at 300 K, implying that Cr_Te crystals are ferromagnetic elemental semiconductors. When B // c // I, the maximum negative MR is -27% at T = 20 K and B = 8 T. In the low temperature semiconducting region, Cr_Te crystals show strong discrete scale invariance dominated logarithmic quantum oscillations when the direction of the magnetic field B is parallel to the [100] crystallographic direction and show Landau quantization dominated Shubnikov-de Haas (SdH) oscillations for B // [210] direction, which suggests the broken rotation symmetry of the Fermi pockets in the Cr_Te crystals. The findings of coexistence of multiple quantum oscillations and ferromagnetism in such an elemental quantum material may inspire more study of narrow bandgap semiconductors with ferromagnetism and quantum phenomena.

cond-mat.mtrl-sci↗
arXiv2023

Influence of the vessel wall geometry on the wall-induced migration of red blood cells

The geometry of the blood vessel wall plays a regulatory role on the motion of red blood cells (RBCs). The overall topography of the vessel wall depends on many features, among which the endothelial lining of the endothelial surface layer (ESL) is an important one. The endothelial lining of vessel walls presents a large surface area for exchanging materials between blood and tissues. The ESL plays a critical role in regulating vascular permeability, hindering leukocyte adhesion as well as inhibiting coagulation during inflammation. Changes in the ESL structure are believed to cause vascular hyperpermeability and entrap immune cells during sepsis, which could significantly alter the vessel wall geometry and disturb interactions between RBCs and the vessel wall, including the wall-induced migration of RBCs and the thickening of a cell-free layer. To investigate the influence of the vessel wall geometry particularly changed by the ESL under various pathological conditions, such as sepsis, on the motion of RBCs, we developed two models to represent the ESL using the immersed boundary method in two dimensions. In particular, we used simulations to study how the lift force and drag force on a RBC near the vessel wall vary with different wall thickness, spatial variation, and permeability associated with changes in the vessel wall geometry. We find that the spatial variation of the wall has a significant effect on the wall-induced migration of the RBC for a high permeability, and that the wall-induced migration is significantly inhibited as the vessel diameter is increased.

q-bio.QM↗
arXiv2023

DSPC: Efficiently Answering Shortest Path Counting on Dynamic Graphs

The widespread use of graph data in various applications and the highly dynamic nature of today's networks have made it imperative to analyze structural trends in dynamic graphs on a continual basis. The shortest path is a fundamental concept in graph analysis and recent research shows that counting the number of shortest paths between two vertices is crucial in applications like potential friend recommendation and betweenness analysis. However, current studies that use hub labeling techniques for real-time shortest path counting are limited by their reliance on a pre-computed index, which cannot tackle frequent updates over dynamic graphs. To address this, we propose a novel approach for maintaining the index in response to changes in the graph structure and develop incremental (IncSPC) and decremental (DecSPC) update algorithms for inserting and deleting vertices/edges, respectively. The main idea of these two algorithms is that we only locate the affected vertices to update the index. Our experiments demonstrate that our dynamic algorithms are up to four orders of magnitude faster processing for incremental updates and up to three orders of magnitude faster processing for hybrid updates than reconstruction.

cs.DB↗
arXiv2023

The hadronic equation of state of HESS J1731-347 from the relativistic mean-field model with tensor coupling

A recent report has identified a central compact object (CCO) within the supernova remnant HESS J1731-347, with a mass and radius of $M=0.77^{+0.20}_{-0.17}M{\odot}$ and $R=10.4^{+0.86}_{-0.78}$ km, respectively. To investigate this light compact star, a density-dependent relativistic mean-field (DDRMF) model, specifically the DDVT model, has been employed. The DDVT model incorporates tensor couplings of vector mesons, which {can} successfully describe the properties of finite nuclei, such as charge radius, binding energy, and spin-orbit splitting. The introduction of tensor coupling reduces the influence of scalar mesons and generates a softer equation of state (EOS) in the outer core of the neutron star. Moreover, it has been found that the crust segment plays a crucial role in reproducing the mass-radius relation of HESS J1731-347, indicating a preference for a soft crust EOS. By manipulating the coupling strength of the isovector meson in the DDVT parameter set, a reasonable hadronic EOS has been obtained, satisfying the constraints from the gravitational-wave signal GW170817, the simultaneous mass-radius measurements from the NICER collaboration, and the properties of finite nuclei. Notably, the mass-radius relations derived from this hadronic EOS also accurately describe the observables of HESS J1731-347. Therefore, based on our estimation, the CCO in HESS J1731-347 may represent the lightest known neutron star.

nucl-th↗
arXiv2023

AoM: Detecting Aspect-oriented Information for Multimodal Aspect-Based Sentiment Analysis

Multimodal aspect-based sentiment analysis (MABSA) aims to extract aspects from text-image pairs and recognize their sentiments. Existing methods make great efforts to align the whole image to corresponding aspects. However, different regions of the image may relate to different aspects in the same sentence, and coarsely establishing image-aspect alignment will introduce noise to aspect-based sentiment analysis (i.e., visual noise). Besides, the sentiment of a specific aspect can also be interfered by descriptions of other aspects (i.e., textual noise). Considering the aforementioned noises, this paper proposes an Aspect-oriented Method (AoM) to detect aspect-relevant semantic and sentiment information. Specifically, an aspect-aware attention module is designed to simultaneously select textual tokens and image blocks that are semantically related to the aspects. To accurately aggregate sentiment information, we explicitly introduce sentiment embedding into AoM, and use a graph convolutional network to model the vision-text and text-text interaction. Extensive experiments demonstrate the superiority of AoM to existing methods. The source code is publicly released at https://github.com/SilyRab/AoM.

cs.CL↗
arXiv2023

Bidirectional Transformer Reranker for Grammatical Error Correction

Pre-trained seq2seq models have achieved state-of-the-art results in the grammatical error correction task. However, these models still suffer from a prediction bias due to their unidirectional decoding. Thus, we propose a bidirectional Transformer reranker (BTR), that re-estimates the probability of each candidate sentence generated by the pre-trained seq2seq model. The BTR preserves the seq2seq-style Transformer architecture but utilizes a BERT-style self-attention mechanism in the decoder to compute the probability of each target token by using masked language modeling to capture bidirectional representations from the target context. For guiding the reranking, the BTR adopts negative sampling in the objective function to minimize the unlikelihood. During inference, the BTR gives final results after comparing the reranked top-1 results with the original ones by an acceptance threshold. Experimental results show that, in reranking candidates from a pre-trained seq2seq model, T5-base, the BTR on top of T5-base could yield 65.47 and 71.27 F0.5 scores on the CoNLL-14 and BEA test sets, respectively, and yield 59.52 GLEU score on the JFLEG corpus, with improvements of 0.36, 0.76 and 0.48 points compared with the original T5-base. Furthermore, when reranking candidates from T5-large, the BTR on top of T5-base improved the original T5-large by 0.26 points on the BEA test set.

cs.CL↗
arXiv2023

A Study of Generative Large Language Model for Medical Research and Healthcare

There is enormous enthusiasm and concerns in using large language models (LLMs) in healthcare, yet current assumptions are all based on general-purpose LLMs such as ChatGPT. This study develops a clinical generative LLM, GatorTronGPT, using 277 billion words of mixed clinical and English text with a GPT-3 architecture of 20 billion parameters. GatorTronGPT improves biomedical natural language processing for medical research. Synthetic NLP models trained using GatorTronGPT generated text outperform NLP models trained using real-world clinical text. Physicians Turing test using 1 (worst) to 9 (best) scale shows that there is no significant difference in linguistic readability (p = 0.22; 6.57 of GatorTronGPT compared with 6.93 of human) and clinical relevance (p = 0.91; 7.0 of GatorTronGPT compared with 6.97 of human) and that physicians cannot differentiate them (p < 0.001). This study provides insights on the opportunities and challenges of LLMs for medical research and healthcare.

cs.CL↗
arXiv2023

From Alignment to Entailment: A Unified Textual Entailment Framework for Entity Alignment

Entity Alignment (EA) aims to find the equivalent entities between two Knowledge Graphs (KGs). Existing methods usually encode the triples of entities as embeddings and learn to align the embeddings, which prevents the direct interaction between the original information of the cross-KG entities. Moreover, they encode the relational triples and attribute triples of an entity in heterogeneous embedding spaces, which prevents them from helping each other. In this paper, we transform both triples into unified textual sequences, and model the EA task as a bi-directional textual entailment task between the sequences of cross-KG entities. Specifically, we feed the sequences of two entities simultaneously into a pre-trained language model (PLM) and propose two kinds of PLM-based entity aligners that model the entailment probability between sequences as the similarity between entities. Our approach captures the unified correlation pattern of two kinds of information between entities, and explicitly models the fine-grained interaction between original entity information. The experiments on five cross-lingual EA datasets show that our approach outperforms the state-of-the-art EA methods and enables the mutual enhancement of the heterogeneous information. Codes are available at https://github.com/OreOZhao/TEA.

cs.CL↗
arXiv2023

A Mountain-Shaped Single-Stage Network for Accurate Image Restoration

Image restoration is the task of aiming to obtain a high-quality image from a corrupt input image, such as deblurring and deraining. In image restoration, it is typically necessary to maintain a complex balance between spatial details and contextual information. Although a multi-stage network can optimally balance these competing goals and achieve significant performance, this also increases the system's complexity. In this paper, we propose a mountain-shaped single-stage design base on a simple U-Net architecture, which removes or replaces unnecessary nonlinear activation functions to achieve the above balance with low system complexity. Specifically, we propose a feature fusion middleware (FFM) mechanism as an information exchange component between the encoder-decoder architectural levels. It seamlessly integrates upper-layer information into the adjacent lower layer, sequentially down to the lowest layer. Finally, all information is fused into the original image resolution manipulation level. This preserves spatial details and integrates contextual information, ensuring high-quality image restoration. In addition, we propose a multi-head attention middle block (MHAMB) as a bridge between the encoder and decoder to capture more global information and surpass the limitations of the receptive field of CNNs. Extensive experiments demonstrate that our approach, named as M3SNet, outperforms previous state-of-the-art models while using less than half the computational costs, for several image restoration tasks, such as image deraining and deblurring.

cs.CV↗
arXiv2023

Temporal and Heterogeneous Graph Neural Network for Financial Time Series Prediction

The price movement prediction of stock market has been a classical yet challenging problem, with the attention of both economists and computer scientists. In recent years, graph neural network has significantly improved the prediction performance by employing deep learning on company relations. However, existing relation graphs are usually constructed by handcraft human labeling or nature language processing, which are suffering from heavy resource requirement and low accuracy. Besides, they cannot effectively response to the dynamic changes in relation graphs. Therefore, in this paper, we propose a temporal and heterogeneous graph neural network-based (THGNN) approach to learn the dynamic relations among price movements in financial time series. In particular, we first generate the company relation graph for each trading day according to their historic price. Then we leverage a transformer encoder to encode the price movement information into temporal representations. Afterward, we propose a heterogeneous graph attention network to jointly optimize the embeddings of the financial time series data by transformer encoder and infer the probability of target movements. Finally, we conduct extensive experiments on the stock market in the United States and China. The results demonstrate the effectiveness and superior performance of our proposed methods compared with state-of-the-art baselines. Moreover, we also deploy the proposed THGNN in a real-world quantitative algorithm trading system, the accumulated portfolio return obtained by our method significantly outperforms other baselines.

q-fin.ST↗
arXiv2023

Nonparametric model for the equations of state of neutron star from deep neural network

It is of great interest to understand the equation of state (EOS) of the neutron star (NS), whose core includes highly dense matter. However, there are large uncertainties in the theoretical predictions for the EOS of NS. It is useful to develop a new framework, which is flexible enough to consider the systematic error in theoretical predictions and to use them as a best guess at the same time. We employ a deep neural network to perform a non-parametric fit of the EOS of NS using currently available data. In this framework, the Gaussian process is applied to represent the EOSs and the training set data required to close physical solutions. Our model is constructed under the assumption that the true EOS of NS is a perturbation of the relativistic mean-field model prediction. We fit the EOSs of NS using two different example datasets, which can satisfy the latest constraints from the massive neutron stars, NICER, and the gravitational wave of the binary neutron stars. Given our assumptions, we find that a maximum neutron star mass is $2.38^{+0.15}_{-0.13} M_\odot$ or $2.41^{+0.15}_{-0.14}$ at $95\%$ confidence level from two different example datasets. It implies that the $1.4 M_\odot$ radius is $12.31^{+0.29}_{-0.31}$ km or $12.30^{+0.35}_{-0.37}$ km. These results are consistent with results from previous studies using similar priors. It has demonstrated the recovery of the EOS of NS using a nonparametric model.

nucl-th↗