Maximum-likelihood estimates are approximated using the EM algorithm where we treat mixture membership $delta_ij$ = 1 if $y_ij$ is generated from the zero point mass as latent indicator variables. The log-likelihood in this extended model is $(1-delta_ij) log f_count(y;mu_i,sigma_i^2 )+delta_ij log pi_j(s_j)+(1-delta_ij)log (1-pi_j (sj))$. The responsibilities are defined as $z_ij = pr(delta_ij=1 | data)$.
Vector of size M with the current negative log-likelihoods.
Vector of size M with the previous iterations negative log-likelihoods.
Vector of size M of the relative differences between the previous and current iteration nll.
Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.