Keyphrases
Neural Network
100%
Symmetric Tensor
100%
Two-hidden-layer
100%
Efficient Identification
100%
Activation Function
50%
Deep Neural Network
25%
Numerical Experiments
25%
Parametrization
25%
Training Phase
25%
Identification Method
25%
Structure Identification
25%
Uniform Approximation
25%
Black Box
25%
Network Training
25%
Spectral Norm
25%
Gradient Descent
25%
Hessian
25%
Nonlinear Program
25%
Intrinsic Symmetry
25%
Diagonal Matrix
25%
Finite Difference Approximation
25%
Sample Complexity
25%
Fully Nonlinear
25%
Stable Recovery
25%
Matrix Subspaces
25%
Frobenius
25%
Heuristic Reasoning
25%
Approximate Hessian
25%
Algebraic Manipulation
25%
Nonlinear Layer
25%
Mathematics
Tensor
100%
Neural Network
100%
Approximates
50%
Matrix (Mathematics)
50%
Posteriori
25%
Diagonalmatrix
25%
Uniform Approximation
25%
Finite Difference Method
25%
Deep Neural Network
25%
Numerical Experiment
25%
Black Box
25%
Engineering
Hidden Layer
100%
Symmetric Tensor
100%
Activation Function
50%
Numerical Experiment
25%
Difference Approximation
25%
Network Training
25%
Diagonal Matrix
25%
Black Box
25%
Gradient Descent
25%
Heuristic Reasoning
25%
Deep Neural Network
25%
Parametrization
25%
Computer Science
Layer Neural Network
100%
Symmetric Tensor
100%
Activation Function
50%
Approximation (Algorithm)
25%
Deep Neural Network
25%
Parametrization
25%
Training Phase
25%
Gradient Descent
25%
Algebraic Manipulation
25%
Diagonal Matrix
25%
Difference Approximation
25%