Missing data
Topic of Mixtures and the EM algorithm, part of Statistical learning
Missing data
| Type | Topic |
| Taught in | Mixtures and the EM algorithm |
| Part | Statistical learning |
Values that were meant to be observed and were not. The EM algorithm treats a mixture’s component labels as missing data: with them the complete-data likelihood is easy to maximise, and the E step replaces them by their conditional probabilities.
In Mixtures and the EM algorithm
The display is a fixed point. The weights \widehat\pi_h appear on the right-hand side through \tau_h, which depends on the whole of \widehat\Psi. The equation does identify the missing data: the weights it asks for are the conditional probabilities of the component indicators.
Used in
See also
EM algorithm · Finite mixture model · Complete-data likelihood · E step · Conditional probability · Posterior membership · EM for a normal mixture · Monotonicity of EM