gromo.tools.compute_optimal_added_parameters#

gromo.tools.compute_optimal_added_parameters(matrix_s: Tensor, matrix_n: Tensor, numerical_threshold: float = 1e-15, statistical_threshold: float = 0.001, maximum_added_neurons: int | None = None) → tuple[Tensor, Tensor, Tensor][source]#

Compute the optimal added parameters for a given layer.

Parameters:

  • matrix_s (torch.Tensor in (s, s)) – square matrix S

  • matrix_n (torch.Tensor in (s, t)) – matrix N

  • numerical_threshold (float) – threshold to consider an eigenvalue as zero in the square root of the inverse of S

  • statistical_threshold (float) – threshold to consider an eigenvalue as zero in the SVD of S{-1/2} N

  • maximum_added_neurons (int | None) – maximum number of added neurons, if None all significant neurons are kept

Returns:

optimal added weights alpha, omega and eigenvalues lambda

Return type:

tuple[torch.Tensor, torch.Tensor, torch.Tensor] in (k, s) (t, k) (k,)