- Mixture Models within the classification framework, by Gaussian, t-mixtures, skew distributions and mixtures of factor analyzers; with emphasis on theoretical properties of the robust estimation, using trimming and constraints.
- Economic inequality, focusing on inequality measures, their properties, statistical inference, limit theorems and applications.
- Computational Statistics: algorithms, for the maximum likelihood estimation in presence of incomplete data (as the EM), under constraints (patterned covariance matrices), for robust estimation and for sampling from huge sample spaces, by complete enumeration (Fréchet class for association measures) or by Markov chains (Partial order sets and their linear extensions).