- Random - Mailing list — Dipartimento di Matematica, UniPI

Seminario Bin Yu al dipartimento G. Catelnuovo
by Gustavo Posta 17 Jun '25

17 Jun '25

Lunedì 23 giugno alle 16:00 nella Sala di Consiglio il prof. Bin Yu (Statistics, EECS, Center for Computational Biology, and Simons Institute UC Berkeley) terrà un seminario dal titolo: Sparse dictionary learning and deep learning in practice and theory Abstract: Sparse dictionary learning has a long history and produces wavelet-like filters when fed with natural image patches, corresponding to the V1 primary visual cortex of the human brain. Wavelets as local Fourier Transforms are interpretable in physical sciences and beyond. In this talk, we will first describe adaptive wavelet distillation (AWD) to turn black-box deep learning models interpretable in cosmology and cellular biology problems while improving predictive performance. Then we present theoretical results that, under simple sparse dictionary models, gradient descent in auto-encoder fitting converges to one point on a manifold of global minima, and which minimum depends on the batch size. In particular, we show that when using a small batch-size as in stochastic gradient descent (SGD) a qualitatively different type of â€œfeature selectionâ€ occurs. Gli interessati sono invitati a partecipare. ---------------------------------------------------------------- Gustavo Posta Dipartimento di Matematica Università di Roma "la Sapienza" P.le A. Moro 2, 00185 Roma Italy web: http://www1.mat.uniroma1.it/~posta e-mail: gustavo.posta(a)uniroma1.it phone: +39-06-4991-4969 ----------------------------------------------------------------- -- *Fai crescere le giovani ricercatrici e i giovani ricercatori*** *con il 5 per mille alla Sapienza* Scrivi il codice fiscale dell'Università *80209930587 **Cinque per mille <https://www.uniroma1.it/it/node/23149>*

1 0

OWABI@BayesComp - 19th June 1pm UK time
by Massimiliano Tamborrino 16 Jun '25

16 Jun '25

Dear All, The final OWABI seminar www.warwick.ac.uk/owabi<http://www.warwick.ac.uk/owabi> of the season will feature two selected talks and will take place live from BayesComp https://bayescomp2025.sg/! The talks will take place on Thursday the 19th June at 1pm UK time (note the different date and time!), and will be given by * Filippo Pagani (University of Warwick): Approximate Bayesian Fusion * Maurizio Filippone (KAUST): GANs Secretly Perform Approximate Bayesian Model Selection with abstracts reported below. The talks will be live streamed from the following link: https://monash.zoom.us/j/81050994376?pwd=aF52Ab2oUlNYPO251d4nDw0jXRemia.1<https://www.google.com/url?q=https%3A%2F%2Fmonash.zoom.us%2Fj%2F81050994376…> Meeting ID: 810 5099 4376 Passcode: 137607 The session will run in a hybrid format, taking place live from BayesCom2025 - 8pm-9pm (Singapore time) - and will be located in room LT50 in the conference venue. We look forward to seeing you online or in person! Best, Massimiliano on the behalf of the OWABI Organisers 1st OWABI Talk: 1-1.30pm Uk time Speaker: Filippo Pagani<https://filippopagani.github.io/> (University of Warwick) Title: Approximate Bayesian Fusion Abstract: Bayesian Fusion is a powerful approach that enables distributed inference while maintaining exactness. However, the approach is computationally expensive. In this work, we propose a novel method that incorporates numerical approximations to alleviate the most computationally expensive steps, thereby achieving substantial reductions in runtime. Our approach retains the flexibility to approximate the target posterior distribution to an arbitrary degree of accuracy, and is scalable with respect to both the size of the dataset and the number of computational cores. Our method offers a practical and efficient alternative for large-scale Bayesian inference in distributed environments. 2nd OWABI Talk: 1.30-2pm Uk time Speaker: Maurizio Filippone<https://www.kaust.edu.sa/en/study/faculty/maurizio-filippone> (KAUST) Title: GANs Secretly Perform Approximate Bayesian Model Selection Abstract: Generative Adversarial Networks (GANs) are popular models achieving impressive performance in various generative modeling tasks. In this work, we aim at explaining the undeniable success of GANs by interpreting them as probabilistic generative models. In this view, GANs transform a distribution over latent variables Z into a distribution over inputs X through a function parameterized by a neural network, which is usually referred to as the generator. This probabilistic interpretation enables us to cast the GAN adversarial-style optimization as a proxy for marginal likelihood optimization. More specifically, it is possible to show that marginal likelihood maximization with respect to model parameters is equivalent to the minimization of the Kullback-Leibler (KL) divergence between the true data generating distribution and the one modeled by the GAN. By replacing the KL divergence with other divergences and integral probability metrics we obtain popular variants of GANs such as f-GANs, Wasserstein-GANs, and Maximum Mean Discrepancy (MMD)-GANs. This connection has profound implications because of the desirable properties associated with marginal likelihood optimization, such as (i) lack of overfitting, which explains the success of GANs, and (ii) allowing for model selection, which opens to the possibility of obtaining parsimonious generators through architecture search. ------ Dr. Massimiliano Tamborrino Reader (Associate Professor) and WIHEA Fellow Department of Statistics University of Warwick https://warwick.ac.uk/tamborrino

1 0

Seminars in Statistics at Collegio Carlo Alberto - Steven CAMPBELL (COLUMBIA UNIVERSITY)
by Matteo Giordano 16 Jun '25

16 Jun '25

SEMINARS IN STATISTICS @ COLLEGIO CARLO ALBERTO <https://www.google.com/url?q=https://www.carloalberto.org/events/category/s…> Venerdì 20/06/2025, presso il Collegio Carlo Alberto, in Piazza Arbarello 8, Torino, si terrà il seguente seminario: ------------------------------------------------ 12.00-13.00 Speaker: Steven CAMPBELL (COLUMBIA UNIVERSITY) Title: A Tractable Triple Problem in Sequential Inference: Filtering, Control, and Stopping Abstract: We study a Bayesian sequential inference problem for the drift of a Brownian motion in which the observer not only chooses their stopping and decision rules, but also controls the rate at which information is acquired. This extension introduces a cost for information acquisition, transforming the classical formulation of Shiryaev (1967) into a joint problem of filtering, control, and optimal stopping. Problems of this kind arise in adaptive experimental design and real-time inference across domains such as healthcare, finance, and engineering, where information is costly and decision-making must be timely. Our framework accommodates a host of classification losses and general running costs on the control process. Despite the added complexity, the problem admits a remarkably simple and semi-explicit solution that exhibits structural parallels with the classical setting. This work highlights a rare instance in which a triple problem—combining inference, control, and stopping—remains analytically tractable. Based on joint work with Georgy Gaitsgori and Richard Groenewald. ------------------------------------------------ Sarà possibile seguire il seminario anche in streaming: chiunque volesse collegarsi è pregato di inviare una email entro *mercoledì 18/06/2025* a matteo.giordano(a)unito.it <mailto:matteo.giordano@unito.it> Il webinar è organizzato dalla "de Castro" Statistics Initiative (www.carloalberto.org/stats <http://www.carloalberto.org/stats>) in collaborazione con il Collegio Carlo Alberto. Cordiali saluti, Matteo Giordano Assistant Professor (RTDA) Department of Economics, Social Studies, Applied Mathematics and Statistics (ESOMAS) www.matteogiordano.weebly.com <https://matteogiordano.weebly.com/>

1 0

Seminario Padova - Filippo Quattrocchi
by alekos.cecchin＠unipd.it 15 Jun '25

15 Jun '25

Buongiorno a tutti, Vorremmo segnalarvi che venerdì prossimo (20 Giugno) in aula 2BC30 (Torre Archimede, Università di Padova) ci sarà un seminario per il ciclo di seminari in Probabilità e Finanza di: Filippo Quattrocchi (Institute of Science and Technology Austria (ISTA)) <https://sites.google.com/view/filippoquattrocchi/> https://sites.google.com/view/filippoquattrocchi/ Title: Acceleration-based Optimal Transport Date: June 20, 2025, at 14:30, 2BC30 Abstract: Finding smooth interpolations between probability measures is a problem of broad interest, with applications ranging from physics to biology, from computer graphics to engineering. In this talk, I will discuss a model in which such interpolations are found by minimising an action functional of the acceleration, and present recent results obtained in collaboration with G. Brigati and J. Maas (ISTA), arXiv:2502.15665. The first result is a fluid-dynamic equivalent formulation of the model, previously conjectured by Y. Chen, G. Conforti, and T. T. Georgiou. We complete the formal argument outlined by the aforementioned authors through the introduction of two new mathematical ingredients. The second result is the description of the (hypoelliptic) Riemannian structure induced by the minimal acceleration-based action on the space of probability measures. Vi aspettiamo numerosi! Alberto Chiarini e Alekos Cecchin Sito web del seminario: https://www.math.unipd.it/~chiarini/seminars/

1 0

Barcelona Summer School of Stochastic Analysis and Quantitative Finance
by Bernardo D'Auria 14 Jun '25

14 Jun '25

Dear all, I would like to share the call for the Barcelona Summer School of Stochastic Analysis and Quantitative Finance, 21-25 July 2025 https://www.crm.cat/5th-barcelona-summer-school-of-stochastic-analysis-and-… Registration deadline July 06th, 2025 Regards, Bernardo D'Auria — Associate Professor at University of Padua

1 0

Doctoral Colloquium on Risk Analytics: applications to 3rd session are open
by Maria Elvira Mancino 13 Jun '25

13 Jun '25

CICF is offering 20 fellowships to doctoral students for attending the third session of the *Doctoral Colloquium in Risk Analytics* <https://landing.cafoscarichallengeschool.it/docra/> *Session 3 (14 - 27 September): AI for Risk * 15-23 September: *Stefano Favaro* (U. Torino and Collegio Carlo Alberto),* Predictive Uncertainty in Machine Learning with Conformal Inference* 19 September: Guest Lecture by *Alessandro Armando *(U. Genova and IMT Lucca), *Introduction to Cyber Risk* 22-26 September: *Christa Cuchiero* (U. Vienna), *Concepts of Deep Learning and Applications to Finance and Risk Management* *Aims and scope* The Colloquium is an *honours course *for Doctoral students interested in analytical (both theoretical and applied) methods for the measurement, management and mitigation of risks, organized in four sessions. The span of risks considered ranges from standard economic, ﬁnancial or insurance risks, to emerging ones, such as environmental, climate and cyber risks. The approach is in-depth and multidisciplinary, with lectures, assignments and discussion of the results. The course is organized into four sessions. *Target* We seek *PhD students from Italian or Foreign Universities*, who remain aﬃliated with their Doctoral School. Students may apply for one session at a time, but must attend the whole session. They may get credit for the courses taken within the Colloquium. There are no participation fees; the Colloquium covers accommodation and may reimburse travel expenses. Up to five additional participants, including further Doctoral Students, Postdocs, Junior Faculty, or PhDs working in the industry, may be admitted at their own expense. *Where* Santa Marta <https://www.camplus.it/en/city/venice/camplus-santa-marta/>, Venice *Applications* Apply through the link at the bottom of the webpage <https://landing.cafoscarichallengeschool.it/docra/> Each PhD student will be asked to provide a recommendation letter from her/his PhD coordinator. Acceptance of qualified applicants will be done on a first-in basis. *The organizers*: Elisa Luciano (chair), Mavira Mancino, Marco Corazza, Marco Li Calzi

1 0

Special Issue on Causal Inference in Healthcare
by Azzimonti Laura 13 Jun '25

13 Jun '25

Dear all, We are pleased to announce the special issue of the International Journal of Approximate Reasoning titled “Causal Inference in Healthcare”. The deadline for manuscript submission is November 5th, 2025. The widespread availability of real-world data from electronic health records, databases, registries, and digital health tools offers unique opportunities for advancing healthcare. However, extracting actionable insights requires grounded methods that account for bias, confounders, and the complexity of medical and biomedical systems. Causal inference provides a rigorous and robust framework for modelling cause-and-effect relationships, addressing biases and confounding factors and evaluating treatment effects, ultimately enabling personalised decision-making. This special issue seeks contributions exploring innovative applications, methodological advancements, and interdisciplinary perspectives on causality in healthcare, with a focus on utilising evidence from diverse data sources to address critical medical challenges. Topics of interest include, but are not limited to: - Personalised treatment effect estimation. - Causal inference in randomised clinical trials. - Bias identification and mitigation in healthcare studies. - Counterfactual reasoning for clinical prediction. - Counterfactual fairness in healthcare. - Mediation analysis in medical research. Guest editors: - Laura Azzimonti, IDSIA USI-SUPSI, Lugano, Switzerland (laura.azzimonti(a)supsi.ch<mailto:laura.azzimonti@supsi.ch>) - Alessandro Antonucci, IDSIA USI-SUPSI, Lugano, Switzerland (alessandro.antonucci(a)supsi.ch<mailto:alessandro.antonucci@supsi.ch>) Important Dates: - Submission Open Date: May 05, 2025 - Submission Deadline: November 05, 2025 - Notification of Acceptance: March 05, 2026 Manuscripts must be submitted via the International Journal of Approximate Reasoning online submission system (Editorial Manager®). Please select the article type “VSI: Causal Inference in Healthcare” when submitting your manuscript online. Please refer to the Guide for authors to prepare your manuscript. All submitted papers under this call will undergo the journal's standard review process. For any further information, the authors may contact the Guest Editors. We warmly encourage you to submit your work or share this call with colleagues who might be interested. For further details, please visit https://www.sciencedirect.com/special-issue/321586/causal-inference-in-heal… Thank you for your attention. Best regards, Laura Azzimonti Guest Editor ––––––––––––––––––––––––––––––––––––––––––––– Scuola universitaria professionale della Svizzera italiana Istituto Dalle Molle di studi sull’intelligenza artificiale IDSIA USI-SUPSI Laura Azzimonti, PhD Senior Researcher and Lecturer Polo universitario Lugano Via la Santa 1 CH – 6962 Lugano-Viganello Tel. +41 (0)58 666 66 44 laura.azzimonti(a)idsia.ch www.idsia.ch

1 0

Seminar Giuseppina GUATTERI
by Antonello Pesce 13 Jun '25

13 Jun '25

Dear all, On JUNE 18th, at the Department of Mathematics of the University of Bologna, room Vitali, and online at this LINK <https://teams.microsoft.com/l/meetup-join/19:meeting_NzYzZTc5ZDgtZjAxYi00ND…>, the following seminar will take place: 11:00-12:00: Giuseppina GUATTERI “MAXIMUM PRINCIPLE FOR OPTIMAL CONTROL PROBLEMS WITH DELAYS IN THE NON CONVEX CASE” as part of the cycle Stochastics and Applications. ABSTRACT: We establish a stochastic maximum principle for controlled stochastic differential equations with delay and control-dependent noise, without convexity assumptions on the control space. The cost functional depends on both present and delayed states, modeled via general finite measures. For measures with square-integrable densities, we employ infinite-dimensional reformulation and BSDE techniques; for general measures, we apply anticipated BSDEs and weak convergence methods. We further analyze the case of delay measures with $L^p$-densities ($p \in (1,2)$), deriving a generalized mild backward equation beyond Hilbertian settings. For any question, please contact the organizers: Elena Bandini (elena.bandini7(a)unibo.it <mailto:elena.bandini7@unibo.it>), Antonello Pesce (antonello.pesce2(a)unibo.it <mailto:antonello.pesce2@unibo.it>)

1 0

Bernoulli Society New Researcher Award 2026 in Probability Theory
by AMORINO, CHIARA 13 Jun '25

13 Jun '25

Dear colleagues, I would like to kindly remind you of the *Bernoulli Society New Researcher Award 2026 in Probability Theory*, which is currently accepting applications. The call has been open for some time, but we have not yet received many applications, so we would really appreciate your help in spreading the word. This award recognizes outstanding contributions from early-career researchers in probability. Each awardee will be invited to present their work at a *special invited session during the 45th Stochastic Processes and their Applications (SPA) conference*, to be held in *Ithaca, New York, USA, June 15–19, 2026*. In addition, each awardee will receive up to *1000 EUR* to help cover travel and other expenses, and their profiles will be featured in *Bernoulli News*. *Eligibility:* - Active researchers in Probability Theory who obtained their PhD on or after *June 30, 2020*. - Candidates must be regular members of the Bernoulli Society. - Extensions may be granted for career breaks, such as parental leave. - The Bernoulli Society strongly encourages applications from women and members of under-represented groups. *Application process:* Candidates should submit their applications through the web form available [here] and send the required documents to the email address indicated in the form. *Deadline:* June 30, 2025 Please feel free to forward this call to potential candidates or anyone who might be interested! Thank you very much for your help. Best regards, Chiara

1 0

DEADLINE EXTENSION: Summer School on "Combinatorial stochastic models and mean field games", 8-12 September 2025, San Miniato, Italy.
by Giulio Bottazzi 12 Jun '25

12 Jun '25

Dear Colleagues, the Centro de Giorgi, Scuola Normale Superiore, and the PhD Programme in Economics of the Scuola Superiore Sant'Anna, are jointly organizing the Ph.D.Summer School on "Combinatorial stochastic models and mean field games". The school is aimed at graduate students in Economics, Finance and related disciplines. Application for last-year master students or young researchers will also be considered, conditional on available founds. The deadline for registration and application for FINANCIAL SUPPORT has been extended to June 23rd. Please check https://indico.sns.it/event/105/ Main lecturers: Daria Ghilli, University of Pavia, and Enrico Scalas, La Sapienza, Rome. Best Regards, Giulio Bottazzi -- Giulio Bottazzi PGP Key ID:BAB0A33F Professor of Economics Faculty of Social Sciences Advanced Calculus for Economics and Finance Theory and Methods https://link.springer.com/book/10.1007/978-3-031-30316-6 Scuola Superiore Sant'Anna 56127 - Pisa, Italy Phone:(+39)-050-883365 email:bottazzi@santannapisa.it Fax: (+39)-050-883344 http://www.sssup.it/~bottazzi The information transmitted is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any review of, re-transmission, dissemination or other use of, or taking of any action in reliance upon, this information by persons or entities other than the intended recipient, is prohibited. If you received this in error, please inform the sender by e-mail and delete the content. Il contenuto e gli allegati di questo messaggio sono strettamente confidenziali e ad uso esclusivo del destinatario. Ne sono vietati la diffusione, l'inoltro e l'uso non autorizzato da parte di utenti diversi dal destinatario. Qualora il presente messaggio Le fosse pervenuto per errore, La preghiamo di distruggerlo e darcene comunicazione a mezzo e-mail.

1 0