expectation: Compute E-step

Description Usage Arguments Value

Description

Compute responsabilities of each cluster on each instance

Usage

1
expectation(priori, gaussParams, vMParams, linearData, angularData)

Arguments

priori

is a vector that represents the a priori probability of an instance belonging to an instace without further information

gaussParams

is a bn.fit object that saves the parameters of the multivariate gaussian Bayesian network

vMParams

is a list saving objects of vonMises Class for each cluster with their parameters mu and kappa

linearData

a dataframe of the variables that represents linear variables

angularData

a dataframe of the variables that represents angular variables

Value

list with two entries, the first one is the matrix of responsabilities the second one is de log-likelihood of the model


sergioluengosanchez/EMS_clustering documentation built on May 31, 2019, 10:37 a.m.