A5100331415- Topic Modeling
- Natural Language Processing Techniques
- Advanced Text Analysis Techniques
- Software Engineering Research
- Advanced Multi-Objective Optimization Algorithms
- Chemical Synthesis and Reactions
- Autonomous Vehicle Technology and Safety
- Quantum Computing Algorithms and Architecture
- Sulfur-Based Synthesis Techniques
- Quantum Information and Cryptography
- Robotics and Sensor-Based Localization
- Geometry and complex manifolds
- Advanced Graph Neural Networks
- Metal-Organic Frameworks: Synthesis and Applications
- Geometric Analysis and Curvature Flows
- Text and Document Classification Technologies
- Biomedical Text Mining and Ontologies
- Metaheuristic Optimization Algorithms Research
- Virtual Reality Applications and Impacts
- Software System Performance and Reliability
- Advanced Computational Techniques and Applications
- Text Readability and Simplification
- Collaboration in agile enterprises
- Optical measurement and interference techniques
- Sentiment Analysis and Opinion Mining
Jiangxi Provincial Cancer Hospital
2025
Tongji University
2022-2024
Shanghai Institute of Pollution Control and Ecological Security
2024
State Key Laboratory of Pollution Control and Resource Reuse
2022-2024
East China Jiaotong University
2023
Chang'an University
2023
Tsinghua University
2022-2023
Shanghai University
2023
International Space Science Institute - Beijing
2023
Shenzhen University
2023
To improve the decision success rate of a multiagent reinforcement learning algorithm in merging high-speed ramps autonomous vehicles, independent proximal policy optimization (IPPO) method is presented. The Markov process (MDP) model for vehicle behavioral making developed. Moreover, state space, reward function, and action space are all designed. An IPPO proposed using parameter-sharing strategies based on algorithm. And further, decision-making driving behavior built. For simulation...
Refining and chemical integration is the major trend in development of world petrochemical industry, showing intensive large-scale development. The accident risks caused by this are complex diverse, pose new challenges to industry safety. In order clarify characteristics risk root contained production process enterprise, avoid reasonably improve overall safety level paper, 159 cases dangerous chemicals China from 2017–2021 were statistically analyzed. A Bayesian network (BN)-based analysis...
The unprecedented performance of large language models (LLMs) necessitates improvements in evaluations. Rather than merely exploring the breadth LLM abilities, we believe meticulous and thoughtful designs are essential to thorough, unbiased, applicable Given importance world knowledge LLMs, construct a Knowledge-oriented Assessment benchmark (KoLA), which carefully design three crucial factors: (1) For ability modeling, mimic human cognition form four-level taxonomy knowledge-related...
The fast and elitist non-dominated sorting genetic algorithm-II (NSGA-II) is currently the most popular multi-objective evolutionary algorithm (MOEA). NSGA-II has been shown to work well for two-objective problems by attaining near-optimal diverse uniformly distributed Pareto solutions. To use powerful optimisation performance of directly conveniently, an classification model presented. In model, a linear equation set constructed according problems. this paper, we introduced solve model....
MeiHan Tong, Bin Xu, Shuai Wang, Meihuan Han, Yixin Cao, Jiangqi Zhu, Siyu Chen, Lei Hou, Juanzi Li. Proceedings of the 2022 Conference North American Chapter Association for Computational Linguistics: Human Language Technologies. 2022.
Existing methods to eliminate the laser speckle noise in quantitative phase imaging always suffer from loss of detailed information and resolution reduction reproduced image. To overcome these problems, this paper proposes a suppression method based on empirical mode decomposition. Our proposed requires only one image without additional equipment avoids complicated process searching optimal processing parameters. In method, we use decomposition highlight high frequency interference Canny...
In this paper, we present an autostereoscopic display system with wide-dynamic-range view zone, high brightness and switchable 2D/3D. For purpose, propose a novel directional backlight inspired by optical splitter. Combined eye tracking technique LED-lit matrix, expand the viewing zone achieve 2D/3D properties. And design of direct illumination-type LED improves significantly. The mathematical analysis logic principle is provided in paper. also, build prototype system, demonstrate...
As a new mode and means of smart manufacturing, cloud manufacturing (SCM) faces great challenges in massive supply demand, dynamic resource collaboration intelligent adaptation. To address the problem, this paper proposes an SCM-oriented supply-demand (S-D) adaptation model for services. In model, collaborative network is established based on properties both their relationships; addition, algorithm deep graph clustering (DGC) aligned sampling (AS) used to divide conquer large domain solve...
Antibiotic-resistant bacteria (ARB) and antibiotic resistance genes (ARGs) are emerging contaminants leading to severe worldwide health problems. Chlorination, a widely used procedure, was extensively explored as one of the main methods remove ARB ARGs in recent years. In this study, enrich analyses chlorination, several comprehensive effects chlorine disinfection system on their conjugative transfer ability were explored. The results presented that low dose (<3-log inactivation rate) had...
The refractive index of the scintillation crystal is a key parameter in design Positron Emission Tomography (PET) devices. However, existing methods to measure always suffer from problems such as complexity optical path adjustment, long time cost, or high equipment consumption. To overcome these problems, we propose method that can obtain with accuracy real time. In this method, employ Mach-Zehnder interferometer generate interference pattern when one beam light passes through crystal, which...
Agricultural productivity is affected by natural disasters, such as drought and flood, which can be assessed remote sensing (RS) technology. The main objective of this research to develop an operational methodology for predicting rice yield that utilized assess the impacts flood disasters on agricultural taking Jiangxi, China example. For purpose, we first calculated Standardized Precipitation Index (SPI) using monthly rainfall data 83 stations from 1960 2020 identify disastrous years their...
An integrated method of CAD, CAE and multi-objective optimization is proposed in this paper. It prototype modeling, FEM analysis procedures. In to design a new product, only inputs outputs need be defined, which resembles black-box. order validate method, manipulator cherry-picker are developed as case study. The objectives extending its working scope decreasing cost under the constraints enough strength, variables geometric dimensions. Multi-objective evolutionary algorithms (MOEAs) applied...
The fast and elitist non-dominated sorting genetic algorithm-II (NSGA-II) is currently the most popular multi-objective evolutionary algorithm (MOEA). NSGA-II has been shown to work well for two-objective problems by attaining near-optimal diverse uniformly distributed Pareto solutions. To use powerful optimisation performance of directly conveniently, an classification model presented. In model, a linear equation set constructed according problems. this paper, we introduced solve model....
Objective Cerebral white matter hyperintensity can lead to cerebral small vessel disease, MRI images in the brain are used assess degree of pathological changes regions. In this paper, we propose a framework for automatic 3D segmentation based on address problems low accuracy and inhomogeneity segmentation. We explored correlation analyses cognitive assessment parameters multiple comparison investigate differences volume among three states, Dementia, MCI NCI. The study between coefficients...
Knowledge Graph has been proven effective in modeling structured information and conceptual knowledge, especially the medical domain. However, lack of high-quality annotated corpora remains a crucial problem for advancing research applications on this task. In order to accelerate domain-specific knowledge graphs domain, we introduce DiaKG, Chinese dataset Diabetes graph, which contains 22,050 entities 6,890 relations total. We implement recent typical methods Named Entity Recognition...
Web and artificial intelligence technologies, especially semantic web knowledge graph (KG), have recently raised significant attention in educational scenarios. Nevertheless, subject-specific KGs for K-12 education still lack sufficiency sustainability from data perspectives. To tackle these issues, we propose EDUKG, a heterogeneous sustainable Educational Knowledge Graph. We first design an interdisciplinary fine-grained ontology uniformly modeling resource education, where define 635...
Open Information Extraction models have shown promising results with sufficient supervision. However, these face a fundamental challenge that the syntactic distribution of training data is partially observable in comparison to real world. In this paper, we propose syntactically robust framework enables be trained on syntactic-abundant based diverse paraphrase generation. To tackle intrinsic problem knowledge deformation paraphrasing, two algorithms semantic similarity matching and tree...