2.
**Statistical Modeling and Inference**

3.
**Signal and Data Compression**

**4.
****Information
Theoretic Methods**

*In 2008-2009* we proposed a number
of new contributions starting from the newest approach to composite hypothesis
testing proposed by Rissanen, which relies on the concept of optimally
distinguishable distributions (ODD). In [671], we derived the ODD detector for
the classical linear model. The work was further extended in a paper published
in *IEEE Transactions on Signal Processing*
(see [249]), where we provided answers to the following problems that have not
been previously investigated in the literature: (i)
the relationship between ODD and the widely used Generalized Likelihood Ratio
Test; (ii) the connection between ODD and the Information Theoretic Criteria
(ITC) applied in model selection.

*NML
models for discrete regression *

__http://www.cs.tut.fi/~tabus/IoanTabusPublicationList.pdf__

__http://www.cs.tut.fi/~bogdand/BD_PublicationList.html__

__http://sp.cs.tut.fi/cgi-bin/cgiwrap/spwww/publication.cgi?aux=giurcaneanu%2C+c__