A5100390341- Machine Learning in Materials Science
- Computational Drug Discovery Methods
- Mass Spectrometry Techniques and Applications
- Advanced machining processes and optimization
- Probabilistic and Robust Engineering Design
- Innovative Microfluidic and Catalytic Techniques Innovation
- Neural Networks and Applications
- Advanced Polymer Synthesis and Characterization
- Network Security and Intrusion Detection
- Corrosion Behavior and Inhibition
- Technology and Security Systems
- X-ray Diffraction in Crystallography
- Structural Health Monitoring Techniques
- Protein Structure and Dynamics
- Human Mobility and Location-Based Analysis
- Material Properties and Applications
- Machine Learning in Bioinformatics
- Receptor Mechanisms and Signaling
- TiO2 Photocatalysis and Solar Cells
- Photochromic and Fluorescence Chemistry
- Data Management and Algorithms
- Polydiacetylene-based materials and applications
- Advanced Multi-Objective Optimization Algorithms
- Fault Detection and Control Systems
- biodegradable polymer synthesis and properties
Beihang University
2022-2024
North University of China
2024
Southwest Petroleum University
2023
Harbin Institute of Technology
2021
Michigan State University
2021
Southern Methodist University
2021
University of Electronic Science and Technology of China
2019
China Pharmaceutical University
2012
Modeling the spatiotemporal relationship (STR) of traffic data is important yet challenging for existing graph networks. These methods usually capture features separately in temporal and spatial dimensions or represent by adopting multiple local spatial-temporal graphs. The first kind method mentioned above difficult to potential temporal-spatial relationships, while other limited long-term feature extraction due its receptive field. To handle these issues, Synchronous Spatio-Temporal grAph...
Monte Carlo (MC) methods are important computational tools for molecular structure optimizations and predictions. When solvent effects explicitly considered, MC become very expensive due to the large degree of freedom associated with water molecules mobile ions. Alternatively implicit-solvent can largely reduce cost by applying a mean field approximation meanwhile maintains atomic detail target molecule. The two most popular models Poisson-Boltzmann (PB) model Generalized Born (GB) in way...
Abstract Numerous challenges arise from metal corrosion, including economic loss, environmental pollution, and safety issues. Epoxy coatings, as a solution to corrosion problems, have been widely used worldwide. The improvement of epoxy coating performance is also major focus research. In this paper, we employed hydrothermal method load UIO‐66 crystals onto two‐dimensional mica substrate, while simultaneously modifying the surface with polyethyleneimine (PEI) enhance compatibility filler in...
Ultraviolet (UV) photochromism is observed from natural light in amorphous TiO2 nanotube arrays (NTAs) for the first time. Surface color of NTAs film would change yellow to dark brown eventually, either under ultraviolet bulb irradiation or basking sunlight. This attributed appearance Ti3 + ions after UV illumination. Furthermore, a radiation cumulative dosimeter designed and fabricated, consisting photochromic colorimetric card, which convert invisible rays into visible changes. device...
Using computer simulation tools such as finite element analysis (FEA) to perform material stress is a common design method in engineering practice. In order model more realistic real‐world systems, models have become complex, and calculation becomes expensive result. The rise of artificial intelligence technologies has made it possible integrate deep learning methods analysis. Herein, FEA software employed obtain large number cases training samples uses fully connected neural network...
Although various knowledge distillation (KD) methods for CNN-based detectors have been proven effective in improving small students, build- ing baselines and recipes DETR-based detec- tors remains a challenge. This paper concentrates on the transformer decoder of explores KD suitable them. However, random order outputs poses challenge as it provides no direct correspondence between pre- dictions teacher student. To this end, we propose MixMatcher that aligns de- coder student, by mixing two...
We present a novel approach for the prediction of crystal material properties that is distinct from computationally complex and expensive density functional theory (DFT)-based calculations. Instead, we utilize an attention-based graph neural network yields high-accuracy predictions. Our employs two attention mechanisms allow message passing on graphs, which in turn enable model to selectively attend pertinent atoms their local environments, thereby improving performance. conduct...
Based on the photo-elastic method, four groups of different ratio light elastic-plastic material were prepared by changing epoxy resin, firming agent, plasticizer mass ratio; and elastic-plasticity stress-strain curve its elastic modulus E Poisson's obtained through stretch experiment.The rail models made 1:5, computation analyzes stress distribution head heavy carried static examination.The stripe chart change rule during load unload process was observed.