ML-adjoint: Learn the adjoint source directly for full-waveform inversion using machine learning

Engineering

• Markov Decision Process 100%
• Full-Waveform Inversion 100%
• Optimisation Problem 50%
• Measured Data 50%
• Jacobian 50%
• Neural Network Architecture 50%
• Demonstrates 50%
• Measured Output 50%
• Objective Function 50%
• Backpropagation 50%

Computer Science

• Machine Learning 100%
• Adjoint Source 100%
• Markov Decision Process 28%
• Optimization Problem 14%
• Jacobian Matrix 14%
• Residual Data 14%
• Fast Convergence 14%
• Backpropagation 14%
• Neural Network Architecture 14%
• Objective Function 14%

Keyphrases

• Markov Decision Process 100%
• Residual Matrix 50%
• Policy Networks 50%