I have fitted a Gaussian mixture model to my data. This Gaussian mixture model is the combination of two Gaussian distributions. I call the first Gaussian distribution state 1 and the second Gaussian distribution state 2. I now want to classify each observations to one of the two states. How could I do this? Do I need the Virterbri algorithm, the Backward-Forward algorithm or some different kind of algorithm?

  • $\begingroup$ Do you have a sequence of observations? Or are each observations independent? $\endgroup$ – jonnor Apr 10 '19 at 10:11
  • $\begingroup$ Yes, I have a sequence of stock returns, which I wish to classify as bull/bear market. They are of course independent between each other, but depend on one of the two mixtures. $\endgroup$ – Cardinal Apr 10 '19 at 10:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.