learn_graph_sigrep {spectralGraphTopology} | R Documentation |
Learn graphs from a smooth signal representation approach This function learns a graph from a observed data matrix using the method proposed by Dong (2016).
Description
Learn graphs from a smooth signal representation approach
This function learns a graph from a observed data matrix using the method proposed by Dong (2016).
Usage
learn_graph_sigrep(
X,
alpha = 0.001,
beta = 0.5,
maxiter = 1000,
ftol = 1e-04,
verbose = TRUE
)
Arguments
X |
a p-by-n data matrix, where p is the number of nodes and n is the number of observations |
alpha |
hyperparameter that controls the importance of the Dirichlet energy penalty |
beta |
hyperparameter that controls the importance of the L2-norm regularization |
maxiter |
maximum number of iterations |
ftol |
relative error on the objective function to be used as the stopping criteria |
verbose |
if TRUE, then a progress bar will be displayed in the console. Default is TRUE |
Value
A list containing the following items
laplacian |
estimated Laplacian Matrix |
Y |
a smoothed approximation of the data matrix X |
convergence |
whether or not the algorithm has converged within the tolerance and max number of iterations |
obj_fun |
objective function value at every iteration, in case record_objective = TRUE |
References
X. Dong, D. Thanou, P. Frossard and P. Vandergheynst, "Learning Laplacian Matrix in Smooth Graph Signal Representations," in IEEE Transactions on Signal Processing, vol. 64, no. 23, pp. 6160-6173, Dec.1, 2016.