A Monte Carlo EM Algorithm for the Estimation of a Logistic Auto-logistic Model with Missing Data
This paper proposes an algorithm for the estimation of the parameters of a Logistic Auto-logistic Model when some values of the target variable are missing at random but the auxiliary information is known for the same areas. First, we derive a Monte Carlo EM algorithm in the setup of maximum pseudo-likelihood estimation; given the analytical intractability of the conditional expectation of the complete pseudo-likelihood function, we implement the E-step by means of Monte Carlo simulation. Second, we give an example using a simulated dataset. Finally, a comparison with the standard non-missing data case shows that the algorithm gives consistent results.
|Date of creation:||2008|
|Date of revision:|
|Contact details of provider:|| Postal: |
Web page: http://www.unitn.it/deco
More information through EDIRC
When requesting a correction, please mention this item's handle: RePEc:trn:utwpde:0801. See general information about how to correct material in RePEc.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Luciano Andreozzi)
If references are entirely missing, you can add them using this form.