E.M. Algorithm: A toll for analysis in pressence of missing value
The expectation maximization algorithm is an broadly applicable approach to the iterative computation of maximum likelihood (ML) estimate useful in variety of incomplete data problem, where algorithm such as Newton-Raphson methods turn out to be a more complicated. On each iteration of the algorithm there are two steps: Expectation step (E-step) and maximization step (M-step). Name due to Dempster Laird and Rubin (1977). For a detailed study on EM algorithm one can go through Lachlan and Krishnan (2008).
# R program to find MLE using EM Algorithm: Multinomial Example x1=125 r+1 # This will gives the number of iteration
References 1. Dempster AP, Laird NM and Rubin DB ,1977, Maximum likelihood from incomplete data via the EM algorithm (with discussion). J Roy Stat Soc B 39:1–38 2. McLachlan GJ, Krishnan T, 2008, The EM Algorithm and Extensions, 2nd Edition, Wiley Series in Probability and Statistics
|