Re: EM ALGORITHM

The EM algorithm is known as Expectation Maximization. It can be used
for Gaussian Mixture Models and even for incomplete data sets. It is
primarily used to estimation of missing values.

-h


mlimber wrote:
> sabbathian@xxxxxxxxx wrote:
> > mlimber wrote:
> > > sabbathian@xxxxxxxxx wrote:
> > > > DOES ANYONE HAVE OR KNOW WHERE I CAN FIND A GOOD AND EASY
> > > > TUTORIAL FOR
> > > > THE EM ALGORITHM WITH EXAMPLES??
> > > >
> > > > THANKS!
> > >
> > > Which EM algorithm are you refering to, and why are you yelling at us?
> > >
> > > Cheers! --M
> > I am sorry about the caps.
> >
> > **Please note, that this is not a homework question.
> > I am looking for an EM algorithm that can be used for missing
> > features/data/values?
> >
> > E.g.
> > The following is a hypothetical example:
> >
> > assume x to be a missing feature/data/value
> >
> >
> > f1 f2 f3 f4 f5
> > f6 f7 f8
> > first point = [0.33 0.045 0.454 1.556 x 0.8 x
> > 1.2]
> > sec point = [0.23 x 0.980 0.64 2.34 x 0.7
> > 1.0]
> > third point= [x 0.2 1.200 0.98 8.2 1.2
> > 1.67 1.3]
> > .
> > .
> > .
> > n-th point=[0.35 0.023 0.444 0.78 x 1.11 0.89
> > 0.9]
> >
> > hence my total dataset is n by 8. The 8 columns each represent a
> > reading of perhaps a sensor. Each sensor may be reading a different
> > feature value. (e.g. gas value, chemical index)
> >
> >
> > How can the EM be used? Furthermore, the problem becomes more
> > complicated if each of the datapoints can be classified as one of m
> > classes( class1 , class2 ,.... ,classm). So we do not know the
> > distribution (i.e. we do not know which class each of the points belong
> > to)
> >
> > A good tutorial or explanation on how to approach the problem will be
> > helpful. Also note, that there can be any number of features and points
> > and are not restricted to 8 features as shown above.
>
> I still don't know what "EM" stands for in this context (element
> matching? ergodic morphology? end mean? elephant march?). Also, I
> suspect your question has less to do with image processing proper and
> more with pattern matching and classification. If so, you'd probably be
> better served in another newsgroup. (If you were concerned about the
> extraction of features rather than their classification, you'd be in
> the right place, however.)
>
> Cheers! --M

.

Relevant Pages

  • Re: EM ALGORITHM
    ... I am looking for an EM algorithm that can be used for missing ... Each sensor may be reading a different ... that there can be any number of features and points ...
    (sci.image.processing)
  • Re: EM ALGORITHM
    ... The EM algorithm is known as Expectation Maximization. ... that there can be any number of features and points ... > more with pattern matching and classification. ...
    (sci.image.processing)
  • Re: EM ALGORITHM
    ... I am guessing you are looking for the Expectation Maximization Algorithm? ... that there can be any number of features and points ... more with pattern matching and classification. ...
    (sci.image.processing)
  • Re: EM ALGORITHM
    ... >> Which EM algorithm are you refering to, and why are you yelling at us? ... that there can be any number of features and points ... more with pattern matching and classification. ...
    (sci.image.processing)
  • Re: Help on choosing a valid pattern: composite or not?
    ... when you have a set of features x1,x2,..,xN, a set of classes ... Once weights in all nodes have stabilized, ... >> I moved that into the training and classification contexts. ...
    (comp.object)