20 Giu 2024 11:00

Theory and Data in Model Building: from Volterra to Machine Learning

Aula 2A, Edificio Delta, Campus Scientifico via Torino, Mestre

Seminario con il Prof. Angelo Vulpiani, Sapienza Università di Roma.

Evento organizzato dal Prof. Domenico De Fazio nell'ambito di Engineering Physics Colloquia.

It is not an exaggeration to say that models are unavoidable in scientific practice and that it is impossible to have real science without them. Even top theories (e.g. electromagnetism or quantum mechanics) are nothing but sophisticated models. The talk is devoted to ”true” models which do not pretend to offer a general description, in particular in the cases where the top theories are not very efficient and can be classified into 4 large classes: I- Oversimplified models; II- Models by analogy; III- Large scale models; IV- Models from data. In class I there is the celebrated Lorenz model; the Lotka-Volterra system is in class II, and it is at the origin of biomathematics. Among the models in class III, we have the effective equations used, e.g., in meteorology and engineering, where only ”relevant variables” are taken into account. In class IV we find the most interesting (and difficult) problem: the building of models just from data without a reference theoretical framework.

* V. Volterra, Variazioni e fluttuazioni del numero d’ individui in specie animali conviventi, Mem. R. Accad. Naz. dei Lincei 2, 31 (1926)
* J.G. Charney, R. Fjortoft and J. von Neumann, Numerical integration of the barotropic vorticity equation Tellus 2, 237 (1950)
* E.N. Lorenz Deterministic nonperiodic flows, J. Atmos. Sci. 20, 130 (1963)
* A. Renyi Dialogues on Mathematics, Holden-Day: San Francisco, CA, USA, 1967.
* M. Baldovin, F.Cecconi, M. Cencini, A. Puglisi and A. Vulpiani, The Role of Data in Model Building and Prediction: A Survey Through Examples Entropy, 20, 807 (2018)

Sarà possibile seguire il seminario anche da remoto.
Gli organizzatori offriranno coffee & cookies ai partecipanti.


L'evento si terrà in inglese


Dipartimento di Scienze Molecolari e Nanosistemi, Università Ca’ Foscari Venezia


Locandina 1489 KB

Cerca in agenda