Convergence properties of the EM algorithm in the mixture model with missing data
Nur Karomah Dwidayati, Zaenuri Zaenuri · Journal of Physics Conference Series · 2018
The research aims toinvestigate the convergence properties of the EM Algorithm in the mixture model with missing data.We have the key property of the EM algorithm is that the incomplete data log-likelihood increases after each iteration of the algorithm. So, the EM sequence of likelihoods must converge if the likelihood is bounded above