A5102918834- Stochastic Gradient Optimization Techniques
- Advanced Graph Neural Networks
- Advanced Neural Network Applications
- Privacy-Preserving Technologies in Data
- Machine Learning and Data Classification
- Graph Theory and Algorithms
- Parallel Computing and Optimization Techniques
- Bone Metabolism and Diseases
- Advanced Data Storage Technologies
- Cryptography and Data Security
- Cloud Computing and Resource Management
- Adversarial Robustness in Machine Learning
- Machine Learning and Algorithms
- Domain Adaptation and Few-Shot Learning
- Bone health and treatments
- Neural Networks and Applications
- Topic Modeling
- Extracellular vesicles in disease
- Complex Network Analysis Techniques
- Imbalanced Data Classification Techniques
- Data Quality and Management
- Recommender Systems and Techniques
- Data Management and Algorithms
- Refrigeration and Air Conditioning Technologies
- Metaheuristic Optimization Algorithms Research
Wuhan University
2022-2025
Second Affiliated Hospital of Nantong University
2021-2024
Renmin Hospital of Wuhan University
2022-2024
Affiliated Hospital of Nantong University
2021-2024
Nantong University
2021-2024
University of Sheffield
2024
Qufu Normal University
2024
Northwest Normal University
2023-2024
Jinan University
2024
Taiyuan University of Technology
2023
We study distributed machine learning in heterogeneous environments this work. first conduct a systematic of existing systems running stochastic gradient descent; we find that, although these work well homogeneous environments, they can suffer performance degradation, sometimes up to 10x, where stragglers are common because their synchronization protocols cannot fit setting. Our contribution is heterogeneity-aware algorithm that uses constant rate schedule for updates before adding them the...
Abstract Traditional fraud detection approaches often use linking entities, such as device, email, and address, to identify fraudulent transactions users. However, methods continue evolve escalate, the fraudsters can fabricate involved entities thus hide their real intent. To make more robust, we incorporate user behaviors in pipeline consider biometric characteristics that are difficult forge. In this work, conduct a detailed study of how behavior data help prevent activity e-commerce. We...
To address the challenge of explosive big data, distributed machine learning (ML) has drawn interests many researchers. Since ML algorithms trained by stochastic gradient descent (SGD) involve communicating gradients through network, it is important to compress transferred gradient. A category low-precision can significantly reduce size gradients, at expense some precision loss. However, existing methods are not suitable for cases where sparse and nonuniformly distributed. In this paper, we...
Abstract Machine Learning (ML) techniques now are ubiquitous tools to extract structural information from data collections. With the increasing volume of data, large-scale ML applications require an efficient implementation accelerate performance. Existing systems parallelize algorithms through either parallelism or model parallelism. But cannot obtain good statistical efficiency due conflicting updates parameters while performance is damaged by global barriers in parallel methods. In this...
Graph Convolutional Network (GCN) is a widely used method for learning from graph-based data. However, it fails to use the unlabeled data its full potential, thereby hindering ability. Given some pseudo labels of data, GCN can benefit this extra supervision. Based on Knowledge Distillation and Ensemble Learning, lots methods teacher-student architecture make better then prediction. these introduce unnecessary training costs high bias student model if teacher's predictions are unreliable....
With the ever-evolving concerns on privacy protection, vertical federated learning (FL), where participants own non-overlapping features for same set of instances, is becoming a heated topic since it enables multiple enterprises to strengthen machine models collaboratively with guarantees. Nevertheless, achieve preservation, FL algorithms involve complicated training routines and time-consuming cryptography operations, leading slow speed.
Behavior alterations in fibroblast-like synoviocytes (FLS) contribute to a pivotal role pathogenesis of rheumatoid arthritis (RA). MiRNAs are closely involved variety pathologic conditions. In the present study, we aimed screen for aberrant expression miRNAs (RA-FLS) further identify altered miR-26a-5p RA-FLS and investigate RA. The was screened by microarray analysis confirmed quantitative real time PCR. effect on proliferation, cell cycle, apoptosis, invasion were studied. verification...
All-reduce is the key communication primitive used in distributed data-parallel training due to high performance homogeneous environment. However, sensitive stragglers and delays as deep learning has been increasingly deployed on heterogeneous environment like cloud. In this paper, we propose analyze a novel variant of all-reduce, called partial-reduce, which provides heterogeneity tolerance by decomposing synchronous all-reduce into parallel-asynchronous partial-reduce operations. We...
Simulating the human mobility and generating large-scale trajectories are of great use in many real-world applications, such as urban planning, epidemic spreading analysis, geographic privacy protect. Although previous works have studied problem trajectory generation, continuity generated has been neglected, which makes these methods useless for practical simulation scenarios. To solve this problem, we propose a novel two-stage generative adversarial framework to generate continuous on road...
Osteoarthritis (OA) is a low‑grade, nonspecific inflammatory disease that affects the entire joint. This condition characterized by synovitis, cartilage erosion, subchondral bone defects, and subpatellar fat pad damage. There mounting evidence demonstrating significance of crosstalk between synovitis destruction in development OA. To comprehensively explore phenotypic alterations destruction, it important to elucidate mechanisms chondrocytes synovial cells. Furthermore, updated iteration...
Graph is ubiquitous in various real-world applications, and many graph processing systems have been developed. Recently, hardware accelerators exploited to speed up systems. However, such hardware-specific are hard migrate across different backends. In this paper, we propose the first tensor-based framework, Tgraph, which can be smoothly deployed run on any powerful (uniformly called XPU) that support Tensor Computation Runtimes (TCRs). TCRs, deep learning frameworks along with their...
Vertical federated learning (VFL) trains model when the features of data samples are scattered over multiple clients. To improve efficiency, a promising approach is to find coreset and use it as smaller training set. However, existing methods produce large there many clients have long running time. address these problems, we propose HaCore for efficient construction in VFL setting. first employs locality sensitive hashing (LSH) map bit signatures locally on clients, then merges local...
In recent years, Graph Neural Networks (GNNs) have achieved remarkable success in many graph mining tasks. However, scaling them to large graphs is challenging due the high computational and storage costs of repeated feature propagation non-linear transformation during training. One commonly employed approach address this challenge model-simplification, which only executes Propagation (P) once pre-processing, Combine (C) these receptive fields different ways then feed into a simple model for...
How to classify and organize the semantic Web services help users find meet their needs quickly accurately is a key issue be solved in era of service-oriented software engineering. This paper makes full use characteristics solid mathematical foundation stable classification efficiency naive bayes method. It proposes service method based on theory bayes. elaborates concrete process how three stages bayesian consideration interface execution capacity. The information gain used determine...
Gradient boosting decision tree (GBDT) is one of the most popular machine learning models widely used in both academia and industry. Although GBDT has been supported by existing systems such as XGBoost, LightGBM, MLlib, system bottleneck appears when dimensionality data becomes high. As a result, we tried to support our industrial partner on datasets dimension up 330K, observed suboptimal performance for all these aforementioned systems. In this paper, ask "Can build scalable training whose...
Graph convolutional networks (GCNs) have been successfully applied in many different real-world tasks. However, most of the existing methods are based on shallow GCN, because multiple layers involve long-distance neighborhood information but lead to over-smoothing problem. Actually, a similar challenge exists depth limitation for primitive neural (CNNs). As multi-layer architecture can increase representation ability we study and learn from recent progress CNN propose Lasagne, novel GCN...