In this paper Object-Oriented Bayesian networks are proposed as a tool to model measurement errors in a categorical variable due to respondent. A mixed measurement error model is presented and an Object-Oriented Bayesian network implementing such a model is introduced. The insertion of evidence represented by the observed value and its propagation throughout the network yields for each unit the probability distribution of the true value given the observed. Two methods are used to predict the individual true value and their performance is evaluated via simulation.
Marella, D., Vicard, P. (2012). OBJECT-ORIENTED BAYESIAN NETWORKS FOR MODELLING THE RESPONDENT MEASUREMENT ERROR. In WORKING PAPERS (DIPARTIMENTO DI ECONOMIA DELL'UNIVERSITÀ DI ROMA TRE)]. ROMA : Dipartimento di Economia, Univ. Roma Tre.
OBJECT-ORIENTED BAYESIAN NETWORKS FOR MODELLING THE RESPONDENT MEASUREMENT ERROR
MARELLA, Daniela;VICARD, Paola
2012-01-01
Abstract
In this paper Object-Oriented Bayesian networks are proposed as a tool to model measurement errors in a categorical variable due to respondent. A mixed measurement error model is presented and an Object-Oriented Bayesian network implementing such a model is introduced. The insertion of evidence represented by the observed value and its propagation throughout the network yields for each unit the probability distribution of the true value given the observed. Two methods are used to predict the individual true value and their performance is evaluated via simulation.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
