EM for a normal mixture
Definition taught in Mixtures and the EM algorithm, part of Statistical learning
EM for a normal mixture
| Type | Definition |
| Taught in | Mixtures and the EM algorithm |
| Part | Statistical learning |
Choose initial values \Psi^{(0)}. For k=0,1,2,\ldots:
- E step: compute \tau_{ih}=\tau_h(\mathbf{x}_i;\Psi^{(k)}) for all i,h.
- M step: compute \pi_h^{(k+1)}, \boldsymbol\mu_h^{(k+1)}, and \Sigma_h^{(k+1)} from the three updates above.
- Stop when the change in \ell(\Psi^{(k)};X_{\mathrm{all}}) is smaller than a threshold.
Used in
- Taught in Mixtures and the EM algorithm, where the derivation and examples are.
- All 7 questions on Mixtures and the EM algorithm
See also
EM algorithm · Monotonicity of EM · Complete-data likelihood · E step · M step · Finite mixture model · Posterior membership · Likelihood of a normal mixture