Menü des aktuellen Bereichs:



Lage Science Park 2

Das Institut für Stochastik befindet sich im Science Park 2, 6. Stockwerk ...  mehr zu Anfahrtsplan (Titel)



Massimiliano Tamborrino

Dr. Massimiliano Tamborrino

Dr. Massimiliano Tamborrino (Zum Team)

S2 0614
Tel.: +43 732 2468 4168

Tab-Navigation zu zugehörigen Informationen
  1. Statistical inference for (partially/full observed) stochastic processes.
  2. Monte-Carlo methods and simulation.
  3. First passage times.
  4. Point processes and dependence measures.
  5. Stochastic modelling in neuroscience.
  6. Neural network connectivity.
  7. Mathematical modeling of physiological systems.
  8. Mathematical modeling of visual attention.

Research interests

My main research interests are inference for (multivariate partially/fully observed) stochastic processes (involving parametric and non-parametric techniques based on Monte-Carlo methods and simulation), hitting times (aka first passage times) and their applications, and mathematical modeling of neuronal and physiological systems.

In particular, I have been (and still am) working on first passage times of multivariate diffusion processes from a statistic (cf. [12]), probabilistic and numerical point of view (cf. [5], [9], [10]). I am also interested in dependence measures between point processes (assuming that each event is in fact a passage time), with application to neural network connectivity (cf. [1] , [2] ).

One of my favorite topic is the statistical inference from perturbed stochastic processes, i.e. stochastic processes in presence of a stimulus onset which modifies the model parameters (cf. [3], [4], [6], [7], [11], [13]). I started working on this topic during my PhD, and then I developed it thanks to two consecutive 2-years bilateral Projects between Austria and Czech-Republic (Czech PI: Prof. Lansky), for 2015-2016 (PI, 3875€) and 2017-2018. The ongoing project is titled "Perturbed stochastic point processes as a novel tool for neural coding analysis" (PI, 6790 €).

Recently I got interested in multi-timescale adapting threshold models, i.e. diffusion processes in presence of time-varying thresholds, having a jump after each crossing time, yielding non-renewal processes of hitting times. A preliminary step in this direction is represented by [9], where I investigated the first passage time density of a Wiener process in presence of exponentially decaying threshold (no jumps occur). Together with Dr. Kobayashi from Japan, I plan to investigate them from a modelling, probabilistic, statistical and numerical point of view. The model under consideration requires the development of techniques to deal with systems of recursive integral equations, non-renewal point processes and analytical methods for considers systems of MTATMs.

In 2017 and 2018 I have been founded for a 2-years bilateral Project between Austria and France (French PI: Prof. Samson) titled "Statistical inference for multivariate partially observed stochastic processes with application to neuroscience" (PI, 6840€). The model under consideration does not fall into the well-known class of Hidden Markov models, requiring further statistical developments.

Co-authors: Susanne Ditlevsen, Lubomir Kostal, Martin Jacobsen, Petr Lansky, Laura Sacerdote and Cristina Zucca.