An unsupervised latent/output physics-informed convolutional-LSTM network for solving partial differential equations using peridynamic differential operator

Mathematics

• Neural Networks 72%
• Differential operator 63%
• Physics 63%
• Output 57%
• Partial differential equation 56%
• Architecture 51%
• Network Architecture 46%
• Memory Term 40%
• Nonlocal Interactions 21%
• Activation Function 18%
• Training 16%
• Extrapolation 16%
• Benchmark 15%
• Governing equation 14%
• Learning 14%
• Filter 13%
• Derivative 9%
• Demonstrate 9%
• Performance 9%
• Form 6%

Engineering & Materials Science

• Partial differential equations 76%
• Mathematical operators 70%
• Physics 66%
• Neural networks 46%
• Network architecture 36%
• Memory architecture 23%
• Extrapolation 20%
• Convolutional neural networks 17%
• Chemical activation 17%
• Derivatives 15%
• Data storage equipment 11%

Physics & Astronomy

• differential operators 100%
• partial differential equations 75%
• physics 48%
• output 45%
• learning 16%
• education 15%
• extrapolation 14%
• activation 12%
• filters 10%
• performance 6%
• interactions 6%