A5042572266- E-commerce and Technology Innovations
- Advanced Graph Neural Networks
- Collaboration in agile enterprises
- Recommender Systems and Techniques
- Supply Chain and Inventory Management
- Cryptography and Data Security
- Anomaly Detection Techniques and Applications
- Sustainable Supply Chain Management
- Scheduling and Optimization Algorithms
- Topic Modeling
- Network Security and Intrusion Detection
- Product Development and Customization
- Digital Platforms and Economics
- Urban and Freight Transport Logistics
- Advanced Technologies in Various Fields
- Privacy-Preserving Technologies in Data
- Advanced Manufacturing and Logistics Optimization
- Vehicle Routing Optimization Methods
- Advanced Image and Video Retrieval Techniques
- Opportunistic and Delay-Tolerant Networks
- Wireless Sensor Networks and IoT
- Complexity and Algorithms in Graphs
- Stochastic Gradient Optimization Techniques
- Machine Learning and Data Classification
- Machine Learning and Algorithms
China Three Gorges University
2008-2025
Sichuan University
2017-2024
Chengdu University
2023
Tongji University
2011
Chongqing University
2010
In the complex working environment of Internet Things (IoT), there are many differences in work between devices, as well associations and energy constraints. The entire task scheduling system needs to consume a large amount for communication. order describe its relevance constraint relationship, traditional modeling methods need add number this paper, an loss optimization method based on multi-objective fuzzy algorithm is proposed. Based equipment cost time IoT environment, equation...
Automatic segmentation of infected lesions from computed tomography (CT) COVID-19 patients is crucial for accurate diagnosis and follow-up assessment. The remaining challenges are the obvious scale difference between different types similarity normal tissues. This work aims to segment scales lesion boundaries correctly by utilizing multiscale multilevel features. A novel dilated convolutional network (MSDC-Net) proposed against low contrast tissues in CT images. In our MSDC-Net, we propose a...
In recent years, graph-based learning methods have gained significant traction in point-of-interest (POI) recommendation systems due to their strong generalization capabilities. These approaches commonly transform user check-in records into graph-structured data and leverage graph neural networks (GNNs) model the representations of both POIs users. Despite effectiveness, GNNs face inherent limitations message passing, which can impede deep extraction meaningful from structure. To mitigate...
At present, wireless sensor networks (WSNs) play an important role in collecting and processing information smart transportation monitoring. Inevitably, the performance of resource scheduling algorithm directly determines quality service WSNs. In this paper, we present a novel large-scale WSNs based on differential ion coevolution multi-objective decomposition (DIC-MOD) to optimize We first introduce certain number mobile nodes with higher configuration into consider them as relay strengthen...
Students’ mental health has always been the focus of social attention, and prediction can be regarded as a time-series classification task. In this paper, an informer network based on two-stream structure (TSIN) is proposed to calculate interdependence between students’ behaviors trend time cycle, intermediate features are integrated layer by realize gating mechanism. Through experiments real campus environment dataset (STU) open (MTS), it verified that algorithm obtain higher accuracy than...
In order to mitigate the influence of human subjectivity on indicator weights in performance evaluation enterprise collaboration, and explore nonlinear relationship between collaboration influencing factors results, this paper propose a combined model based AHP-EW an improved Elman neural network. Firstly, characteristics among manufacturing enterprises, system for collaborative enterprises is constructed from three dimensions. Moreover, study combines subjective objective weighting methods...
This paper proposes a logistics tracking information management system based on wireless sensor network, using nodes to track and manage information, designs networked achieve remote real‐time of information. The research work in this can mention the accuracy management, which has significant guiding significance security industry. first analyzes business process enterprises, obtains demand analysis system, divides whole into several functional modules according these modules, gives model...
Abstract In recent years, understanding human behavior has become one of the most important topics in field computer vision research. The reason for this growing attention is wide range applications that can benefit from results analysis (HBA) includes a research areas detection motion and action. Datasets created through actions activities make it possible to compare different methods with same input data. Data mining big data approaches are very popular analyzing related be used address...
As the development of Internet Things (IoT) continues, Federated Learning (FL) is gaining popularity as a distributed machine learning framework that does not compromise data privacy each participant. However, held by enterprises and factories in IoT often have different distribution properties (Non-IID), leading to poor results their federated learning. This problem causes clients forget about global knowledge during local training phase then tends slow convergence degrades accuracy. In...
Value chain collaboration management is an effective means for enterprises to reduce costs and increase efficiency enhance competitiveness. Vertical horizontal have received much attention, but the current model combining two weak in terms of task assignment node constraints whole production-distribution process. Therefore, enterprise dynamic alliance, this paper models MVC (multi-value-chain) process optimization needs network other aspects. Then a constructed with lowest total cost as...
Batch process monitoring datasets usually contain missing data, which decreases the performance of data-driven modeling for fault identification and optimal control. Many methods have been proposed to impute data; however, they do not fulfill need data quality, especially in sensor with different types data. We propose a hybrid imputation method batch multi-type In this method, is first classified into five categories based on continuous duration number variables simultaneously. Then, are...
In order to solve large-scale or super traveling salesman problem efficiently and quickly, according the general model of plant growth simulation algorithm, a solution method based on is proposed. By calculating testing real data instances, results show that optimum can be reached efficiency presented algorithm reasonably, proposed has better performance.