### p3 ML

Verfasst:

**13. Jun 2008 14:18**is the maximal likelihood for multinomial distribution quite similar to gaussian distribution?

Fachschaft Informatik

FB Informatik

TU Darmstadt

https://www.fachschaft.informatik.tu-darmstadt.de/forum/

https://www.fachschaft.informatik.tu-darmstadt.de/forum/viewtopic.php?f=292&t=12576

Seite **1** von **1**

Verfasst: **13. Jun 2008 14:18**

is the maximal likelihood for multinomial distribution quite similar to gaussian distribution?

Verfasst: **13. Jun 2008 14:40**

Not really. The E-step of the EM algorithm looks somewhat similar, but otherwise no.

But just to make sure: All the formulas that you need to implement are given on the assignment sheet. You do not need to make any derivations.

But just to make sure: All the formulas that you need to implement are given on the assignment sheet. You do not need to make any derivations.

Verfasst: **13. Jun 2008 14:50**

i'm quite confused. the formular is just for EM algorithm, not for ML...

Verfasst: **13. Jun 2008 18:51**

can't figure it out. better give it up

Verfasst: **13. Jun 2008 21:13**

The EM algorithm that you need to implement is actually doing (approximate) ML estimation.

If you look back at the slides for mixture models, you will notice that there is no closed form solution for exact ML estimation in mixture models. Hence we need to approximate somehow. The EM algorithm allows us to do that, because in every iteration the (incomplete) likelihood either goes up or stays the same.

If you look back at the slides for mixture models, you will notice that there is no closed form solution for exact ML estimation in mixture models. Hence we need to approximate somehow. The EM algorithm allows us to do that, because in every iteration the (incomplete) likelihood either goes up or stays the same.