A reminder of today's seminar:
Monday 1 November 1999 - 4pm, Room 102, Pearson Building
Department of Statistical Science - University College London
Speaker : Paola Sebastiani (Maths, Open University)
Title : Robust Bayesian Classification
Abstract :
The talk describes the Robust Bayes Classifier, which is Naive Bayes
Classifier trained on an incomplete data set using the Robust Bayesian
Estimator. The resulting classifier is quantified by probability intervals
and we provide an algorithm for propagating these probability intervals to
compute bounds on the posterior probability for each class, given a set of
attribute values. Such intervals can then be used for classification by
choosing the class associated with the highest ranked interval. We provide
two scoring methods to rank intervals and a decision theoretic approach to
trade off the risk of an erroneous classification and the choice of not
classifying unequivocally a case. Such decision theoretic approach can also
be used to help one assess the adequacy of the Missing at Random assumption
on a particular database. The work is in collaboration with Marco Ramoni.
Useful Web sites:
The programme of our statistical seminars is available at:
http://www.ucl.ac.uk/Stats/research/journals.html
Map with instructions on how to get to our department:
http://www.ucl.ac.uk/Stats/map.html
Regards,
Kostas Skouras
----------------------------------------------------------------------------
---------------
K. Skouras, Lecturer in Statistics, Department of Statistical Science,
University College London, Gower Street, London WC1E 6BT, United Kingdom.
Tel: 0171-4193652 Fax: 0171-7383 4703
e-mail: [log in to unmask]
----------------------------------------------------------------------------
---------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|