0
$\begingroup$

I used Mplus to perform a latent class analyses/finite mixture analyses with categorical and continuous indicator variables. I obtained a well defined 5 cluster solution. I would like now to predict the cluster probabilities for new subjects (without rerunning the analyses again) but do not know how to get the formula done For example I have got the results for a cluster 5

Cluster 5 and mplus code

   Latent Class 5

Many thanks!

$\endgroup$
0
$\begingroup$

I now used the parameters froim output and constrained all variables in th e model. It does correctly classify new patients but if someone knows how to do it manually I would highly appreciate it - I just would like to better understand it.

$\endgroup$
  • 1
    $\begingroup$ I've never worked with Mplus. The Mplus output posted above was confusing to me in that I couldn't tell if it represented parameter estimates or some other summary statistics about the inputs. Having worked with a competing LC package, Latent Gold, I know that it offers class-specific parameter estimates which are to be used in "manually" estimating and classifying new units. This would be the kind of thing you would need to work around automatic model scoring. $\endgroup$ – Mike Hunter Nov 17 '15 at 14:10

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.