mkkm_mr: Multiple kernel k-means with matrix induced regularization...

In vittoriofortino84/COPS: Clustering algorithms for Omics-driven Patient Stratification

Multiple kernel k-means with matrix induced regularization (Liu et al. 2016)

Description

Multiple kernel k-means with matrix induced regularization (Liu et al. 2016)

Usage

mkkm_mr(
  K_list,
  k,
  lambda,
  tolerance = 1e-06,
  parallel = 0,
  use_mosek = FALSE,
  mosek_verbosity = 0L,
  mkkm_mr_maxiter = 10,
  no_stop = FALSE
)

Arguments

K_list	list of kernel matrices
k	number of clusters
lambda	numeric regularization parameter
tolerance	numeric stopping criterion value
parallel	number of parallel threads used by the quadratic solver
use_mosek	If TRUE, the optimization will be run with Rmosek instead of CVXR.
mosek_verbosity	MOSEK logging parameter.
mkkm_mr_maxiter	maximum number of iterations, usually the algorithm converges soon after 2 iterations
no_stop	If TRUE, always runs mkkm_mr_maxiter iterations

Value

a kernel matrix

vittoriofortino84/COPS documentation built on Jan. 28, 2025, 3:16 p.m.