t_update

a matrix of dimension <code>N</code> (number of observation) <code>x M</code> (number of variables) containing multinomials observations.

matrix of dimension <code>M x H</code>.

Theta

matrix of dimension <code>H x K</code>.

Lambda

vector containing the proportions of each cluster. Must be of dimension <code>K</code>.

internal

Provides a version of the Expectation-Maximization algorithm for
mix-models, reducing the numbers of parameters to estimate using
Non-negative Matrix Factorization methods.
For more explanations, see pre-print of Carel and Alquier (2017) <arXiv:1709.03346>.

t_update: Update of t

Description

Usage

Arguments

Value