pyriemann_qiskit.optimization.distance.weights_logeuclid_to_convex_hull

pyriemann_qiskit.optimization.distance.weights_logeuclid_to_convex_hull(A, B, optimizer=<pyriemann_qiskit.optimization.docplex.ClassicalOptimizer object>)[source]

Weights for Log-Euclidean distance to a convex hull of SPD matrices.

Weights for Log-Euclidean distance between a SPD matrix B and the convex hull of a set of SPD matrices A [1], formulated as a Constraint Programming Model (CPM) [2].

Parameters:
Returns:

weights – Optimized weights for the set of SPD matrices A. Using these weights, the weighted Log-Euclidean mean of set A provides the matrix of the convex hull closest to matrix B.

Return type:

ndarray, shape (n_matrices,)

Notes

Added in version 0.0.4.

Changed in version 0.2.0: Rename from logeucl_dist_convex to weights_logeuclid_to_convex_hull. Add linear constraint on weights (sum = 1).

References

[1]

K. Zhao, A. Wiliem, S. Chen, and B. C. Lovell, ‘Convex Class Model on Symmetric Positive Definite Manifolds’, Image and Vision Computing, 2019.