- Random - Mailing list — Dipartimento di Matematica, UniPI

Summer school "Applied Harmonic Analysis and Machine Learning" - Genoa, 7-11 September 2026
by Ernesto De Vito 09 Jan '26

09 Jan '26

As part of the scientific activity and educational offer of MaLGa - Machine Learning Genoa Center (https://malga.unige.it) we are delighted to announce the summer school "Applied Harmonic Analysis and Machine Learning", which will take place in Genoa on September 7–11, 2026. The intended audience includes graduate students in Mathematics, Physics, Computer Science and Engineering, as well as postdoctoral fellows and young researchers. The lecturers of this sixth edition will be Michael Bronstein (University of Oxford), Philipp Petersen (University of Vienna) and Clarice Poon (University of Warwick). We will also host a one-day workshop with invited speakers and selected contributed talks/posters from the participants. The school will be held in person only. Limited funding is available to cover accommodation expenses. The registration form, as well as further information and updates, can be found on the webpage https://malga.unige.it/education/schools/ahaml2026/. We kindly ask you to circulate this message among all the potentially interested colleagues/participants. We look forward to seeing you in Genoa! The organizers G. S. Alberti N. Aroua F. De Mari E. De Vito S. Sanna M. Santacesaria

1 0

SPASS seminar: Pierfrancesco Beneventano
by Marco Romito 09 Jan '26

09 Jan '26

Dear Colleagues, We would like to invite you to the following SPASS seminars, *Edge of Stochastic Stability: SGD does not train neural networks as you expect it* by Pierfrancesco Beneventano (MIT) on Tuesday 13.1.2026 at 14:00 CET in Saletta Riunioni, Dipartimento di Matematica, UNIPI. A link for online participants will be shared on the website https://sites.google.com/unipi.it/spass Best, M. Romito on behalf of the organizers ------------------- Abstract: Recent findings demonstrate that when training neural networks using full-batch (deterministic) gradient descent with step size η, the largest eigenvalue λ of the Hessian consistently stabilizes around 2/η. These results are surprising and carry significant implications for convergence and generalization. This, however, is not the case for mini-batch optimization algorithms, limiting the broader applicability of the consequences of these findings. We show that mini-batch Stochastic Gradient Descent (SGD) trains in a different regime, which we term Edge of Stochastic Stability (EoSS). In this regime, what stabilizes at 2/η is Batch Sharpness: the expected directional curvature of mini-batch Hessians along their corresponding stochastic gradients. As a consequence, λ---which is generally smaller than Batch Sharpness---is suppressed, aligning with the long-standing empirical observation that smaller batches and larger step sizes favor flatter minima. We further discuss implications for mathematical modeling of SGD trajectories.

1 0

BEE-VOLatility Project final event
by Maria Elvira Mancino 09 Jan '26

09 Jan '26

Dear colleagues, on *January 15th, 2026*, the following workshop will take place *Assessing and Forecasting Climate Change and Its Impact on Pollinators’ Health and Ecosystem Integrity* *HONEY BEE VOLatility Final Event <https://www.unive.it/data/33113/7/110160>* The workshop seeks to explore and discuss a broad range of perspectives on climate and environmental risks, from modelling and forecasting climate change to investigating the impacts of climate change on honeybees’ health. The event will also feature contributions developed as part of the PRIN 2022 PNRR project “Honey BEE VOLatility: An environmental index for assessing climatic risk impact on ecosystem service provision”. *Thursday, 15 January 2026, 10.30-16.30* San Giobbe Economics Campus - Venice School of Management, Polo Rispoli (Cannaregio 873, Venice) *Also online via Zoom* Registration (*free*) before 13/1/2026 at 12.00 : https://forms.gle/Ao3BbAeEjyB8PfB97 This event is part of the dissemination activities developed within the Project “Honey BEE VOLatility: An environmental index for assessing climatic risk impact on ecosystems service provision” prot. nr. P2022AX5TH and received funding from the European Union Next-GenerationEU - National Recovery and Resilience Plan (NRRP) – MISSION 4 COMPONENT 2, INVESTIMENT 1.1 Fondo per il Programma Nazionale di Ricerca e Progetti di Rilevante Interesse Nazionale (PRIN) – CUP N. H53D23009460001 Kind regards Mavira

1 0

Meeting "Stochastic Models in Biomathematics and Applications"
by Antonio DI CRESCENZO 08 Jan '26

08 Jan '26

Dear Colleagues, On behalf of the Organizing Committee, I am pleased to announce the Meeting *Stochastic Models in Biomathematics and Applications,* which will take place at the *University of Salerno *(Department of Mathematics) on *January 20–21, 2026*. Scope of the Meeting is the scientific exchange between scientists working in Probability and Mathematical Statistics, with special focus on Stochastic Models, Biomathematics and Applications. The participation is open to any interested scholar. The Meeting falls within the activities of the Project P2022XSF5H supported by the “European Union – Next Generation EU” through MUR-PRIN 2022 PNRR. *Invited Speakers:* · Nikolai Leonenko – *Cardiff University* · Yuliya Mishura – *University of Kyiv* · Thomas Simon – *University of Lille* · Miguel Ángel Sordo Díaz – *University of *Cádiz · Alfonso Suarez Llorens – *University of *Cádiz Participation is free, but registration is mandatory. *Registration and abstract submission deadline:* *January 15, 2026* Further information is available on the website: https://www.dipmat2.unisa.it/iniziative/smod2026/authors.html We look forward to welcoming you to Salerno! Best regards, Antonio Di Crescenzo *On behalf of the Organizing Committee* ============================================= Antonio Di Crescenzo Dipartimento di Matematica Università degli Studi di Salerno Via Giovanni Paolo II, n. 132 <https://maps.google.com/?q=Via+Giovanni+Paolo+II,+n.+132+%0D%0A84084+Fiscia…> 84084 Fisciano (SA) Italy Tel. +39-089-963349 E-mail(1): adicrescenzo(a)unisa.it E-mail(2): adicresc(a)gmail.com Web: http://www.unisa.it/docenti/antoniodicrescenzo/index =============================================

1 0

Seminari Filippo De Feo e Mattia Martini - Politecnico di Milano - 8 Gennaio 2026
by Luca Scarpa 07 Jan '26

07 Jan '26

Si ricordano i seguenti 2 seminari a tutti gli interessati. SEMINARIO 1: Giovedì 8 Gennaio 2026, ore 11:00. Aula Seminari III Piano, Dipartimento di Matematica, Politecnico di Milano. Speaker: Filippo De Feo, Technische Universität Berlin. Title: Stochastic Optimal Control of Interacting Particle Systems in Hilbert Spaces and Applications in Economics. Abstract: Optimal control of interacting particles governed by stochastic evolution equations in Hilbert spaces is an open area of research. Such systems naturally arise in formulations where each particle is modeled by stochastic partial differential equations, path-dependent stochastic differential equations (such as stochastic delay differential equations or stochastic Volterra integral equations), or partially observed stochastic systems. In this talk we present a limiting theory as the number of particles tends to infinity. We apply the developed theory to problems arising in economics where the particles are modeled by stochastic partial differential equations and stochastic delay differential equations. The talks is based on [F. de Feo, F. Gozzi, A. Swiech, L. Wessels, Stochastic Optimal Control of Interacting Particle Systems in Hilbert Spaces and Applications, arXiv preprint arXiv:2511.21646] Link Zoom: https://polimi-it.zoom.us/j/92373976166?pwd=pvQAMGRkAnZ6WLaEXawDkGmbxa61VV.1 Link Seminario Polimi: https://www.mate.polimi.it/eventi/?id=2663&sezione_di_ricerca=probstat&stri… SEMINARIO 2: Giovedì 8 Gennaio 2026, ore 11:45. Aula Seminari III Piano, Dipartimento di Matematica, Politecnico di Milano. Speaker: Mattia Martini, École polytechnique, Centre de Mathématiques Appliquées, Palaiseau. Title: Regularisation in mean field models via infinite dimensional diffusion. Abstract: This talk aims to show how randomizing the dynamics in mean field models can help regularize the associated partial differential equations on the space of probability measures. A key challenge in this approach lies in constructing a suitable notion of noise on the space of probability measures. To this end, we rely on the Dirichlet–Ferguson diffusion process, as studied by Dello Schiavo [AOP 22]. We first examine the effect of this noise on a system of uncontrolled interacting particles and show that it induces a regularizing effect at the level of the corresponding backward Kolmogorov equation. We then analyze a mean field control problem driven by this noise and prove that the associated Hamilton–Jacobi equation admits a unique solution in an appropriate functional space, even when the coefficients have limited regularity. The talk is based on a joint work with F. Delarue (Nice) and G. Sodini (Vienna). Link Zoom: https://polimi-it.zoom.us/j/92373976166?pwd=pvQAMGRkAnZ6WLaEXawDkGmbxa61VV.1 Link Seminario Polimi: https://www.mate.polimi.it/eventi/?id=2664&sezione_di_ricerca=probstat&stri… ----- Prof. Luca Scarpa, PhD Associate Professor in Probability Department of Mathematics Politecnico di Milano Via E. Bonardi 9 20133 Milano, Italy e.mail: luca.scarpa(a)polimi.it<mailto:luca.scarpa@polimi.it> url: https://sites.google.com/view/lucascarpa

1 0

seminario Felix Liebrich in Bicocca
by Fabio Bellini 07 Jan '26

07 Jan '26

Il giorno lunedì 12 gennaio alle ore 17 presso la aula seminari 4096 al quarto piano dell'edificio U7 della Università di Milano-Bicocca, il dott. Felix-Benedikt Liebrich della Università di Amsterdam terrà un seminario dal titolo Lambda-quantiles: A closer look at the new kid on the block Abstract: Lambda-VaR has initially been suggested by Frittelli et al. (2014) as a generalisation of classical Value-at-Risk. Recent years have seen a surge in work on these functionals (or slightly more generally: on Lambda-quantiles), both in financial and insurance applications and from a decision-theoretic perspective. This talk focuses on Lambda-quantiles of univariate probability distributions which generalise classical quantiles by replacing the defining constant threshold in the unit interval with a function Lambda mapping the real line to the unit interval. Working under minimal assumptions, we analyse how the choice of Lambda determines mathematical properties of the resulting functional. In particular, we examine aggregation behaviour, continuity, identifiability of the functional parameter Lambda, and the preservation of finer properties of classical quantiles such as ordinal covariance. We conclude by discussing interpretational issues in financial applications. Based on joint work with Fabio Bellini. Tutti gli interessati sono invitati a partecipare. Fabio Bellini °°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°° Prof. Fabio Bellini Dipartimento di Statistica e Metodi Quantitativi (DISMEQ) Università degli Studi di Milano-Bicocca Edificio U7, IV piano, stanza 4055 Via Bicocca degli Arcimboldi 8, 20126 Milano Webex: https://unimib.webex.com/meet/fabio.bellini Orcid: https://orcid.org/0000-0002-3597-808X Scholar: https://scholar.google.it/citations?user=P61L8P4AAAAJ&hl=it °°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°° °°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°

1 0

Seminario Daniela Witten presso Politecnico di Milano
by Laura Maria Sangalli 07 Jan '26

07 Jan '26

Si avvisa che domani, 08/01/2026, alle ore 14:30 presso il Politecnico di Milano, Dipartimento di Matematica, Aula Saleri, nell’ambito delle attività del MOX, si svolgerà il seguente seminario: Daniela Witten Università di Washington Titolo: Data thinning and beyond Abstract: Contemporary data analysis pipelines often involve the use and reuse of data. For instance, a scientist may explore a dataset to select an interesting hypothesis, and then wish to test this hypothesis with the same data. From a statistical perspective, this double use of data is highly problematic: it induces dependence between the hypothesis generation and testing stages, which complicates inference. Failure to account for this dependence renders classical inference techniques invalid. I will present "data thinning", a set of strategies for obtaining independent training and test sets so that the former can be used to select a hypothesis, and the latter to test it. Data thinning enables valid selective inference in settings for which no solutions were previously available. However, it is also restrictive, in the sense that it requires strong distributional assumptions. Therefore, I will also present strategies inspired by data thinning that enable valid post-selection inference without such assumptions. Il link per seguire il seminario online sarà reso disponibile pochi minuti prima dell’avvio del seminario al seguente indirizzo https://mox.polimi.it/mox-colloquia-seminars-list/mox-seminars/?id_evento=2… Tutti gli interessati sono invitati a partecipare. Cordialmente, Laura Sangalli —— Laura Maria Sangalli MOX - Dipartimento di Matematica Politecnico di Milano Piazza Leonardo da Vinci 32 20133 Milano - Italy (+39) 02 2399 4554 laura.sangalli(a)polimi.it https://sangalli.faculty.polimi.it

1 0

Deadline for Registration Approaching - DeFi & Crypto Workshop
by Daniele Maria DI NOSSE 07 Jan '26

07 Jan '26

Dear all, This is a reminder that the deadline for registration (*2026-01-07*) for the *DeFi & Crypto Workshop*, which will be held at Scuola Normale Superiore (Pisa, Italy) *from 26/01/2026 to 28/01/2026*, is approaching. The workshop explores key policy, technological, and market dynamics of the crypto ecosystem, with a focus on DeFi mechanisms, market microstructure, risk, and price formation. The program features nine invited talks by leading academics and industry experts, complemented by contributed sessions highlighting current trends and future challenges in DeFi. For more information please refer to the official website <https://sns-defi.github.io/>. Best regards, DeFi & Crypto Workshop Organizing Committee

1 0

MLCC Summer School 2026 in Genoa, Italy | Application open!
by Ernesto De Vito 23 Dec '25

23 Dec '25

******************************************************************************* Summer School @MaLGa - Machine Learning Crash Course 2026 ******************************************************************************* We are pleased to announce the Summer School Machine Learning Crash Course (MLCC) 2026 https://malga.unige.it/education/schools/mlcc2026/. The course is organized by MaLGa - the Machine Learning Genoa Center, as part of the ELLIS Genoa unit activities, and is taught by Nicoletta Noceti, Cesare Molinari and Lorenzo Rosasco. MLCC provides an introduction to fundamental concepts and algorithms of Machine Learning. The course presents the statistical learning theory framework with empirical risk minimization and regularization as guiding principles for algorithm design. Linear and nonlinear models will be discussed, including kernel methods and neural networks. The course also covers optimization aspects, discussing and analyzing gradient descent methods and backpropagation. Further advanced topics will be covered in a half-day workshop. Invited speakers will be announced soon! The course is suitable for undergraduate/graduate students, as well as professionals. -------------------------------------------------------------------------------------------------- Format: The course will be held from June 15 to June 19, 2026 at the DIBRIS/DIMA, Via Dodecaneso 35, 16146 Genoa, Italy. It will consist of 20 hours of classes and labs + a workshop; lessons will be in-person only. Participants will receive a certificate of attendance upon completing the course. -------------------------------------------------------------------------------------------------- Applications: Submit your application here: https://forms.gle/FbCMwzFCX9KSLm6j8 Application deadline: March 1, 2026. The maximum number of participants is 120. Selection will be based on CV evaluation. The instructors and the organizers Nathalie Baxs, Giulia Casu, Cesare Molinari, Nicoletta Noceti, Lorenzo Rosasco ******************************************************************************* Ernesto De Vito DIMA - Dipartimento di Matematica MUR Excellence Department 2023-2027 MaLGa - Machine Learning Genoa center Via Dodecaneso 35 16146 Genova Italy e-mail: ernesto.devito(a)unige.it

1 0

[Call for Talks & Papers] MAF2026 – Mathematical and Statistical Methods for Actuarial Sciences and Finance
by Marco CORAZZA 23 Dec '25

23 Dec '25

We are pleased to announce the international conference *MAF2026 — Mathematical and Statistical Methods for Actuarial Sciences and Finance* *https://www.crm.cat/maf-2026/ <https://www.crm.cat/maf-2026/>* - *Dates*: 8–10 April 2026. - *Organization*: Department of Actuarial, Financial and Economic Mathematics, University of Barcelona. - *Venue*: Centre de Recerca Matemàtica, and UAB Bellaterra campus, Barcelona, Spain. *Scope / Objective*Promote interdisciplinary interaction among mathematicians, statisticians, professionals in actuarial science and finance; present new theoretical/methodological developments and applications in actuarial mathematics, financial mathematics, statistical finance/insurance methods, stochastic processes, quantitative risk management, financial/actuarial data modeling, calibration, model validation, regulatory risk. *Call for contributed talks*- *Abstract length*: 300–600 words, including title. - *Abstract format*: PDF, titled "surname_name.pdf". - *Abstract submission via*: Conference registration system on the website. - *Deadline for abstract submission*: 12 January 2026. - *Notification of acceptance*: By 2 February 2026. *Conference proceedings* - *Short-paper**s*: authors are invited to submit full paper (max 10 pages); after peer review, papers published in a Springer volume. - *Manuscript submission for proceedings*: Online, via email to congres.maf2026.papers(a)ub.edu. - *Paper submission deadline*: 20 April 2026. - *Notification of acceptance*: 25 May 2026. - *Final revised version due*: 1 June 2026. *Registration*- *Early registration fee*: €300 (deadline: 23 February 2026). - *Late registration fee*: €350 (deadline: 27 March 2026). - *Registration includes*: Coffee breaks, welcome reception, lunch tickets, social dinner. *Contact for questions*: Ms. Núria Hernández - nhernandez(a)crm.cat. On behalf of the Scientific Committee, Marco Corazza - Ca' Foscari University of Venice, Venezia, Italy Marilena Sibillo - University of Salerno, Salerno, Italy

1 0