CMeansConfig

Properties

k

integerdefault 2≥ 1

Number of clusters.

max_iter

integerdefault 100≥ 1

Maximum number of iterations.

dist

stringdefault "euclidean"

Distance measure.

one of"euclidean""manhattan"

method

stringdefault "cmeans"

"cmeans" = fuzzy c-means; "ufcl" = on-line update.

one of"cmeans""ufcl"

m

numberdefault 2> 1

Degree of fuzzification.

rate_par

number | nulldefault null≥ 0≤ 1

Learning rate for the online (ufcl) variant.

weightstunable

numberdefault 1

Case weights: a single number applied to every case, or a per-case vector.

control

object

Control parameters passed to the clustering backend.

Relationships