A5100435527- Sparse and Compressive Sensing Techniques
- Distributed Sensor Networks and Detection Algorithms
- Blind Source Separation Techniques
- Statistical Methods and Inference
- Stochastic Gradient Optimization Techniques
- Advanced Image and Video Retrieval Techniques
- Fluid Dynamics Simulations and Interactions
- Advanced Bandit Algorithms Research
- Multimodal Machine Learning Applications
- Ship Hydrodynamics and Maneuverability
- Multi-Criteria Decision Making
- Machine Learning and Algorithms
- Microwave Imaging and Scattering Analysis
- Domain Adaptation and Few-Shot Learning
- Computational Fluid Dynamics and Aerodynamics
- Image and Signal Denoising Methods
- Fluid Dynamics and Heat Transfer
- Advanced Statistical Methods and Models
- Interconnection Networks and Systems
- Photoacoustic and Ultrasonic Imaging
- Recommender Systems and Techniques
- Risk and Portfolio Optimization
- Advanced Algorithms and Applications
- Geoscience and Mining Technology
- Evaluation Methods in Various Fields
Bellevue Hospital Center
2018-2025
Cognitive Research (United States)
2020-2024
University of Science and Technology Liaoning
2024
China University of Geosciences
2024
Shenzhen Children's Hospital
2019-2024
State Administration of Cultural Heritage
2023
Dalian Polytechnic University
2014-2023
Taiyuan University of Technology
2023
Wuchang University of Technology
2023
Wuhan University
2023
There has been considerable interest in random projections, an approximate algorithm for estimating distances between pairs of points a high-dimensional vector space. Let A Rn x D be our n dimensions. The method multiplies by matrix R RD k, reducing the dimensions down to just k speeding up computation. typically consists entries standard normal N(0,1). It is well known that projections preserve pairwise (in expectation). Achlioptas proposed sparse replacing N(0,1) with -1,0,1 probabilities...
To study the influence of information on behavior stock markets, a common strategy in previous studies has been to concatenate features various sources into one compound feature vector, procedure that makes it more difficult distinguish effects different sources. We maintain capturing intrinsic relations among multiple is important for predicting trends. The challenge lies modeling complex space and types studying this market behavior. For purpose, we introduce tensor-based framework predict...
Sparse recovery aims to reconstruct sparse signals from compressed linear measurements. In this paper, we propose a algorithm called multiple orthogonal least squares (MOLS), which extends the well-known (OLS) by allowing <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$L$</tex-math> </inline-formula> indices be selected per iteration. Owing its ability catch support in each selection, MOLS often converges...
Traditionally, the task of cross-modal retrieval is tackled through joint embedding. However, global matching used in embedding methods often fails to effectively describe matchings between local regions image and words text. Hence they may not be effective capturing relevance text image. In this work, we propose a heterogeneous attention network (HAN) for efficient retrieval. The proposed HAN represents an by set bounding box features sentence word features. determined set-to-set To enhance...
Atopic dermatitis (AD) is the most common skin disorder in infancy. However, diagnosis and definite significance of infantile AD remains a debated issue.To analyse phenotypes infancy, to establish diagnostic criteria estimate prevalence this condition China.This multicentric study, which 12 locations were chosen from different metropolitan areas China. Following careful complete history-taking examination, was made severity based on SCORAD index determined by local experienced...
Deep neural networks (DNNs) have become state-of-the-art in many application domains. The increasing complexity and cost for building these models demand means protecting their intellectual property (IP). This paper presents a novel DNN framework that optimizes the robustness of embedded watermarks. Our method is originated from fault attacks. Different prior end-to-end watermarking approaches, we only modify tiny subset weights to embed watermark, which also facilities better control model...
We study an important variant of the stochastic multi-armed bandit (MAB) problem, which takes penalization into consideration. Instead directly maximizing cumulative expected reward, we need to balance between total reward and fairness level. In this paper, present some new insights MAB formulate problem in framework, where a rigorous penalized regret can be well defined more sophisticated analysis is possible. Under such propose hard-threshold UCB-like algorithm, enjoys many merits...
This R\&D project, initiated by the DOE Nuclear Physics AI-Machine Learning initiative in 2022, leverages AI to address data processing challenges high-energy nuclear experiments (RHIC, LHC, and future EIC). Our focus is on developing a demonstrator for real-time of high-rate streams from sPHENIX experiment tracking detectors. The limitations 15 kHz maximum trigger rate imposed calorimeters can be negated intelligent use streaming technology system. approach efficiently identifies low...
Deep neural network has been adopted as the standard model to predict ads click-through rate (CTR) for commercial online advertising systems. Deploying an industrial scale system requires overcome numerous challenges, e.g., hundreds or thousands of billions input features and also training samples, which under cost budget can cause fundamental issues on storage, communication, speed. In this work, we present Baidu's industrial-scale practices how apply machine learning techniques address...
Hard Thresholding Pursuit (HTP) is an iterative greedy selection procedure for finding sparse solutions of underdetermined linear systems. This method has been shown to have strong theoretical guarantee and impressive numerical performance. In this paper, we generalize HTP from compressive sensing a generic problem setup sparsity-constrained convex optimization. The proposed algorithm iterates between standard gradient descent step hard thresholding with or without debiasing. We prove that...
Intuitionistic fuzzy distance measurement is an effective method to study multi-attribute emergency decision-making (MAEDM) problems. Unfortunately, the traditional intuitionistic cannot accurately reflect difference between membership and non-membership data, where it easy cause information confusion. Therefore, from number (IFN), this paper constructs a model based on cross-entropy comprehensive grey correlation analysis algorithm. For MAEDM problems of completely unknown partially known...
Terahertz imaging has found extensive applications in non-destructive testing, security inspection, and other various fields. Intensive research on terahertz systems been executed to pursue high performance resolution depth of field (DOF). However, the with both large DOF have rarely reported. In this paper, a mesoscopic-sized dielectric bilayer sphere-assisted super-resolution method was proposed simultaneously achieve enormously improved extended DOF. Simulation analyses revealed that...
The method of stable random projections is a useful tool for efficiently computing the lα (0 < α ≤ 2) norms and distances in massive data one pass. Consider matrix A ∈RnxD. If we multiply with projection R ΕR Dxk (k« D),whose entries are i.i.d. samples an α-stable distribution,then projected B = Ax Ε nxkx containsenough information to approximately recover l properties A.