HiDeNN-TD: Reduced-order hierarchical deep learning neural networks

INIS

• decomposition 100%
• tensors 60%
• neural networks 60%
• learning 55%
• finite element method 30%
• accuracy 25%
• solutions 15%
• errors 15%
• degrees of freedom 10%
• levels 5%
• engineering 5%
• cost 5%
• foundations 5%
• efficiency 5%
• comparative evaluations 5%
• size 5%
• reduction 5%
• calculation methods 5%
• increasing 5%
• convergence 5%

Engineering

• Deep Learning 55%
• Finite Element Modeling 30%
• Accuracy 15%
• Degree of Freedom 10%
• Meshes 10%
• High Accuracy 10%
• Converges 5%
• Discretization Error 5%
• Additional Cost 5%
• Reduced Order Model 5%
• Deep Neural Network 5%
• Solution Error 5%
• Foundations 5%
• Efficiency 5%
• High Potential 5%
• Reduction 5%
• Error 5%
• Large Size 5%
• Small Amount 5%

Computer Science

• Accuracy 15%
• Decomposition Method 10%
• Hierarchical Decomposition 10%
• Computation 5%
• Scientific Problem 5%
• Convergence 5%
• Engineering Problem 5%
• Adaptivity 5%
• Discretization Error 5%
• Solution Error 5%
• Hierarchical Order 5%
• Reduced Decomposition 5%