Sandra Fortini

Sandra Fortini is Associate Professor of Statistics in the Department of Decision Sciences, at Bocconi University.

He has previously been Research fellow (1990-1994) and Researcher (1994-1995) at the Institute for the Applications of Mathematics and Informatics of the National Council of Research, and then Assistant Professor (1995-2007) at Bocconi University.

He is Fellow of the Institute of Mathematics Statistics and of the International Society of Bayesian Analysis, of the Italian Statistical Society (SIS), and of the Italian Mathematical Society (UMI).

Sandra Fortini has graduated in Mathematics at Milano University in 1989, and has gained a Master of Science in Applied Stochastic Systems in 1993 at University College of London.

Associate Professor

Associate Professor of Statistics

sandra.fortini@unibocconi.it

Download the cv (325.83 KB)

Research interests

Asymptotic properties of Bayesian and Maching Learning procedures. Predictive inference. Bayesian Nonparametric models.

Selected Publications

Bordino, Alberto; Favaro, Stefano; Fortini, Sandra

30546

PROBABILITY

40027

STOCHASTIC PROCESSES I

40313

INTRODUCTION TO PROBABILITY

40971

PROBABILITY THEORY I

20592

STATISTICS AND PROBABILITY

Past teaching

Sandra Fortini

Non-asymptotic approximations of Gaussian neural networks via second-order Poincaré inequalities

Uncertainty directed factorial clinical trials

Infinitely wide limits for deep Stable neural networks: sub-linear, linear and super-linear activation functions

Prediction-based uncertainty quantification for exchangeable sequences

Infinite-color randomly reinforced urns with dominant colors

Deep Stable neural networks: large-width asymptotics and convergence rates

Infinite-channel deep convolutional Stable neural networks

Approximating the operating characteristics of Bayesian Uncertainty directed trial Designs

Predictive constructions based on measure-valued Pólya urn processes

Large-width functional asymptotics for deep Gaussian neural networks

Stable behaviour of infinitely wide deep neural networks

Quasi‐Bayes properties of a procedure for sequential learning in mixture models

On a notion of partially conditionally identically distributed sequences

Predictive Characterization of Mixtures of Markov Chains

Predictive distribution (de Finetti's view)

Hierarchical reinforced urn processes

Predictive construction of priors in Bayesian nonparametrics

Central Limit Theorem with Exchangeable Summands and Mixtures of Stable Laws as Limits

Recursive equations for the predictive distributions of some determinantal processes

A Poisson Approximation for Coloured Graphs Under Exchangeability

A characterization for mixtures of semi-Markov processes

On mixtures of distributions of Markov chains

On the use of concentration functionin Bayesian robustness

Exchangeability, predictive distributions and parametric models

On parametric models for invariant probability measures

Differential properties of the concentration function.

A central limit problem for partially exchangeable random variables. 1996.

Concentration function and sensitivity to the prior

On defining neighbowrhoods of measures through the concentration function.

Concentration functions and Bayesian robustness

Concentration function and coefficients of divergence for signed measures