A5100424312- Wireless Communication Security Techniques
- Stochastic Gradient Optimization Techniques
- Sparse and Compressive Sensing Techniques
- Advanced Photonic Communication Systems
- Wireless Signal Modulation Classification
- Antenna Design and Analysis
- Statistical Methods and Inference
- Optical Network Technologies
- Microwave Engineering and Waveguides
- Advanced Wireless Communication Technologies
- Advanced Bandit Algorithms Research
- Photonic and Optical Devices
- Advanced MIMO Systems Optimization
- Machine Learning and Algorithms
- Antenna Design and Optimization
- Advanced battery technologies research
- Internet Traffic Analysis and Secure E-voting
- Topic Modeling
- Advanced Optical Network Technologies
- PAPR reduction in OFDM
- Neural Networks and Applications
- Radiomics and Machine Learning in Medical Imaging
- Millimeter-Wave Propagation and Modeling
- Advanced Measurement and Detection Methods
- Optical Wireless Communication Technologies
University of Electronic Science and Technology of China
2018-2025
Nanjing University of Aeronautics and Astronautics
2025
Nanchang Institute of Technology
2024
Beijing University of Posts and Telecommunications
2020-2024
Institute of Coal Chemistry
2024
University of Chinese Academy of Sciences
2024
Chinese Academy of Sciences
2018-2024
Nanchang Institute of Science & Technology
2024
Beijing Technology and Business University
2024
Sichuan University
2024
We study two communication-efficient algorithms for distributed statistical optimization on large-scale data. The first algorithm is an averaging method that distributes the N data samples evenly to m machines, performs separate minimization each subset, and then averages estimates. provide a sharp analysis of this average mixture algorithm, showing under reasonable set conditions, combined parameter achieves mean-squared error decays as O(N <sup...
Globally, the incidence of cervical cancer ranks fourth among female malignant tumors, seriously threatening physical and mental health women. Early detection early treatment can greatly reduce mortality "cytology /HPV test, colposcopy biopsy" is main method for clinical diagnosis cancer. The progress medical technology has significantly improved cancer, but due to various factors, there are still many cases missed misdiagnosis. In recent years, artificial intelligence developed rapidly in...
Heteroatom doping has been demonstrated to be an effective strategy improve the catalytic activity of carbon materials. Herein, heteroatom-doped nanocarbons were found environmental protection cocatalysts for promoting Fenton oxidation. Nitrogen-doped reduced graphene oxide (N-rGO) exhibited better than sulfur-, boron-, and phosphorus-doped rGO enhancing Unlike classical electron sacrificial agents, H2O2 was employed as donor enhance oxidation during catalysis N-rGO. Electrochemical analysis...
We analyze two communication-efficient algorithms for distributed optimization in statistical settings involving large-scale data sets. The first algorithm is a standard averaging method that distributes the N samples evenly to m machines, performs separate minimization on each subset, and then averages estimates. provide sharp analysis of this average mixture algorithm, showing under reasonable set conditions, combined parameter achieves mean-squared error (MSE) decays as O(N-1 +(N/m)-2)....
Background Lymph node metastasis (LNM) is the principal risk factor for poor outcomes in early‐stage cervical cancer. Radiomics may offer a noninvasive way predicting stage of LNM. Purpose To evaluate radiomic signature LN involvement based on sagittal T 1 contrast‐enhanced (CE) and 2 MRI sequences. Study Type Retrospective. Population In all, 143 patients were randomly divided into two primary validation cohorts with 100 cohort 43 cohort. Field Strength/Sequence CE sequences at 3T....
Abstract Sulfide oxidation reaction (SOR) is one central step of electrochemical desulfurization and sulfur‐based batteries. However, the performance sulfur batteries has been severely hindered by passivation. Here, a discovery sulfophobic phenomenon electrocatalysts having weak interaction to species reported. A self‐cleaning NiS 2 electrode developed avoid long‐perplexing passivation issue solid during SOR. Furthermore, sulfur‐vacancies are engineered into lattice synthesize v‐NiS for...
We provide two fundamental results on the population (infinite-sample) likelihood function of Gaussian mixture models with $M \geq 3$ components. Our first main result shows that has bad local maxima even in special case equally-weighted mixtures well-separated and spherical Gaussians. prove log-likelihood value these can be arbitrarily worse than any global optimum, thereby resolving an open question Srebro (2007). second EM algorithm (or a first-order variant it) random initialization will...
We study the Stochastic Gradient Langevin Dynamics (SGLD) algorithm for non-convex optimization. The performs stochastic gradient descent, where in each step it injects appropriately scaled Gaussian noise to update. analyze algorithm's hitting time an arbitrary subset of parameter space. Two results follow from our general theory: First, we prove that empirical risk minimization, if is point-wise close (smooth) population risk, then achieves approximate local minimum polynomial time,...
Background Locally advanced rectal cancer (LARC) patient stratification by clinicoradiologic factors may yield variable results. Therefore, more efficient prognostic biomarkers are needed for improved risk of LARC patients, personalized treatment, and prognostication. Purpose/Hypothesis To compare the ability a radiomic signature to predict disease‐free survival (DFS) with that model in individual patients LARC. Study Type Retrospective study. Population In all, 108 consecutive (allocated...
We propose a robust transceiver design for covert integrated sensing and communications (ISAC) system with imperfect channel state information (CSI). Considering both bounded probabilistic CSI error models, we formulate worst-case outage-constrained optimization problems of joint beamforming radar waveform to balance the performance multiple targets while ensuring covertness system. The are challenging due non-convexity arising from semi-infinite constraints (SICs) coupled variables. In an...
In the rapidly advancing landscape of 6G, characterized by ultra-high-speed wideband transmission in millimeter-wave and terahertz bands, our paper addresses pivotal task enhancing physical layer security (PLS) within near-field communications. We introduce true-time delayer (TTD)-incorporated analog beamfocusing techniques designed to address interplay between propagation beamsplit, an uncharted domain existing literature. Our approach maximizing secrecy rates involves formulating...
We investigate an uplink MIMO-OFDM localization scenario where a legitimate base station (BS) aims to localize user equipment (UE) using pilot signals transmitted by the UE, while unauthorized BS attempts UE eavesdropping on these pilots, posing risk UE's location privacy. To enhance performance protecting privacy, we formulate optimization problem regarding beamformers at aiming minimize Cram\'er-Rao bound (CRB) for constraining CRB above threshold. A penalty dual decomposition framework is...
We investigate the performance tradeoff between \textit{bistatic positioning (BP)} and \textit{monostatic sensing (MS)} in a multi-input multi-output orthogonal frequency division multiplexing scenario. derive Cram\'er-Rao bounds (CRBs) for BP at user equipment MS base station. To balance these objectives, we propose multi-objective optimization framework that optimizes beamformers using weighted-sum CRB approach, ensuring weak Pareto boundary. also introduce two mismatch-minimizing...
With the diversification of online social platforms, news dissemination has become increasingly complex, heterogeneous, and multimodal, making fake detection task more challenging crucial. Previous works mainly focus on obtaining relationships via retweets, limiting accurate when real cascades are inaccessible. Given proven assessment spreading influence events, this paper proposes a method called HML (Complex Heterogeneous Multimodal Fake News Detection Latent Network Inference)....
We investigate a multi-low Earth orbit (LEO) satellite system that simultaneously provides positioning and communication services to terrestrial user terminals. To address the challenges of channel estimation in LEO systems, we propose novel two-timescale positioning-aided framework, exploiting distinct variation rates position-related parameters gains inherent channels. Using misspecified Cramer-Rao bound (MCRB) theory, systematically analyze performance under practical imperfections, such...