[Forum SIS] Seminar Michael Fop

[Marco Alfo']

Apologies for cross-posting


---- SEMINAR (PhD School in Statistical Sciences) ----
Thursday 10 may 2018, 14.00-15.00 room 34,
IVth floor, Statistical Sciences building, Sapienza Universita' di Roma

*Speaker: Michael Fop, University College Dublin.Variable Selection for
Latent Class Analysis.*
*Abstract:* Latent class analysis is a popular approach for model-based
clustering of multivariate categorical data. Typically, all the variables
are considered in fitting the model, but often only a subset of them
contains the useful information about the group structure of the data. This
talk outlines the main features of the variable selection problem and
presents a method for selection of the relevant variables in latent class
analysis. The method is based on the comparison of two different models via
an approximation to their Bayes factor and allows the discarding of
non-informative and redundant variables. An application to the
identification of most relevant clinical criteria for low back pain
diagnosis is presented.



Marco Alfo'
Dipartimento di Scienze Statistiche
Sapienza Università di Roma
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.stat.unipg.it/pipermail/sis/attachments/20180502/d98255c7/attachment-0001.html>