Prof. Chandola, great presentation. Would you please remind me of the statistic(s) to use to evaluate goodness of fit when working in LCA? I remember you mentioned that Chi -squared was not to be used, and that the Log-Likelihood should decrease as the number of classes increases. Any other way of determine whether the number of classes is optimal or should be greater/smaller?
Wonderfully elucidative. Thank you for your presentation!
Very good explanation!! Thanks!
excellent! Thanks very much
Thank you! Very informative.
Informative, but as many have said, the volume is very low.
Prof. Chandola, great presentation. Would you please remind me of the statistic(s) to use to evaluate goodness of fit when working in LCA? I remember you mentioned that Chi -squared was not to be used, and that the Log-Likelihood should decrease as the number of classes increases. Any other way of determine whether the number of classes is optimal or should be greater/smaller?
Very helpful!! Could you suggest some bibliographical references about LCA methodology?
Thanks!
very much helpful, thanks a lot.
The volume is very low...
the volume in noaudible
latent variable models, latent class analysis,