Random

random@lists.dm.unipi.it

30 participants
5305 discussions

Seminario De Palma a Pisa
by Dario Trevisan 27 Mar '19

27 Mar '19

Cari tutti, giovedì 4 aprile, alle ore 14:30, presso la Sala Riunioni del Dipartimento di Matematica (primo piano) dell'Università degli studi di Pisa si terrà il seguente seminario: (Dear all, on Thursday, April 4th, 2:30 PM, the following talk will take place in Sala Riunioni, Dipartimento di Matematica, University of Pisa:) SPEAKER: Giacomo De Palma (University of Copenhagen) TITLE: Random deep neural networks are biased towards simple functions ABSTRACT: We prove that the binary classifiers of bit strings generated by random wide deep neural networks are biased towards simple functions. The simplicity is captured by the following two properties. For any given input bit string, the average Hamming distance of the closest input bit string with a different classification is at least sqrt(n/(2πlog(n))), where n is the length of the string. Moreover, if the bits of the initial string are flipped randomly, the average number of flips required to change the classification grows linearly with n. On the contrary, for a uniformly random binary classifier, the average Hamming distance of the closest input bit string with a different classification is one, and the average number of random flips required to change the classification is two. Our proof is based on the recent breakthrough in machine learning stating that random deep neural networks behave as Gaussian processes. Our results are confirmed by numerical experiments on deep ne ural networks with two hidden layers, and settle the conjecture stating that random deep neural networks are biased towards simple functions. The conjecture that random deep neural networks are biased towards simple functions was proposed and numerically explored in [Valle Pérez et al., arXiv:1805.08522] to explain the unreasonably good generalization properties of deep learning algorithms. By providing a precise characterization of the form of this bias towards simplicity, our results open the way to a rigorous proof of the generalization properties of deep learning algorithms in real-world scenarios. -- Dario Trevisan, Università degli Studi di Pisa, Dipartimento di Matematica, Largo Bruno Pontecorvo 5, 56127 - Pisa (PI) Italy telephone: (+39) 050 2213832 mobile: (+39) 331 2899761 Skype: dario-trevisan e-mail: dario.trevisan AT unipi.it webpage: http://people.dm.unipi.it/trevisan/

1 0

Annonce Postdoc -- Paris Dauphine
by Niccolo Torri 26 Mar '19

26 Mar '19

Ricevo e inoltro. Buona giornata, Niccolò Torri Post-Doc Université Paris-Est Créteil, LAMA, Bat. P2, 61 Av. Général de Gaulle, 94010 Créteil, Bureau P2-232 http://perso.math.u-pem.fr/torri.niccolo/index.html %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Dear all, We are pleased to announce the opening of a 12-month post-doctoral position as part of the ANR MALIN (Interactive Random Walks) for the 2019/20 academic year, with flexible start and end dates. The themes of the ANR are - Gaussian Free Field (GFF), Branching Random Walks (BRW), and Random Interlacements - Random walks in random environments (RWRE) - Self-Interacting random walks (SIRW) - Particle systems, forest fires and percolation but all applications in the field of probability/statistical physics related to these themes will also be carefully considered. The candidate will be affiliated to CEREMADE, Université Paris-Dauphine (*). The file must consist of the following documents: - Curriculum vitae - List of publications - Research statement - 2 letters of recommendation Applications should be sent as soon as possible, and in any case before 3 April 2019, via the CNRS website : https://emploi.cnrs.fr/Offres/CDD/UMR7534-JULPOI-001/Default.aspx With kind regards, Pierre Tarrès (*) PS: Here is the list of Parisian members of the ANR MALIN: Elie Aidekon, Sorbonne Université Anne-Laure Basdevant, Université Paris Nanterre Eric Brunet, Sorbonne Université Bernard Derrida, Sorbonne Université/Ecole Normale Supérieure Paris Laure Dumaz, Université Paris Dauphine Nathalie Eisenbaum, Université Paris Descartes Nathanaël Enriquez, Université Paris-Sud Anna Erschler, Ecole Normale Supérieure Paris Gabriel Faraud, Université Paris Nanterre Nicolas Fournier, Sorbonne Université Yueyun Hu, Université Paris 13 Cyrille Lucas, Université Paris Diderot Titus Lupu, Sorbonne Université Satya Majumdar, Université Paris-Sud Bastien Mallein, Université Paris 13 Julien Poisat, Université Paris Dauphine Olivier Raimond, Université Paris Nanterre Grégory Schehr, Université Paris-Sud Zhan Shi, Sorbonne Université François Simenhaus, Université Paris Dauphine Damien Simon, Sorbonne Université Arvind Singh, Université Paris-Sud Pierre Tarrès, Université Paris Dauphine (en détachement à NYU Shanghaï) Laurent Tournier, Université Paris 13 Olivier Zindy, Sorbonne Université

1 0

Assegno di ricerca all'Aquila
by davide gabrielli 25 Mar '19

25 Mar '19

Si segnala l'assegno di ricerca dal titolo "Comportamento su larga scala di particelle in interazione" Responsabile Prof. Anna De Masi, presso L' UNIVERSITA DEGLI STUDI DELLAQUILA DIPARTIMENTO DI INGEGNERIA E SCIENZE DELLINFORMAZIONE E MATEMATICA. Il bando e' reperibile all'indirizzo http://www.univaq.it/section.php?id=766 ed e' anche in allegato al presente messaggio. Scadenza 09/04/2019 cordiali saluti Davide Gabrielli

1 0

extended deadline: 2019 Bocconi Summer School on Random Graphs
by Antonio Lijoi 25 Mar '19

25 Mar '19

************************************************************* Applications are still open for the 2019 edition of the Bocconi Summer School in Advanced Statistics & Probability on RANDOM GRAPHS AND COMPLEX NETWORKS (July 8-19, 2019) INFO: http://bocconi2019.lakecomoschool.org EXTENDED DEADLINE FOR APPLICATIONS: April 2, 2019 ************************************************************* --- Dai il tuo 5x1000 alla Bocconi. Aiuta gli studenti meritevoli a costruire il proprio futuro. CF 80024610158. www.unibocconi.it/5x1000 Please note that the above message is addressed only to individuals filing Italian income tax returns. 5x1000 is a percentage of Italian personal income tax that taxpayers can allocate to Universities, scientific research and non profit organizations. ---

1 0

An afternoon in Probability: Torino meets Leeds
by Elvira Di Nardo 24 Mar '19

24 Mar '19

%---------------------------------------------------------------------------% An afternoon in Probability: Torino meets Leeds Torino, April 16, 2019 %----------------------------------------------------------------------------% Room 09 - ESOMAS Department Corso Unione Sovietica 218/bis, 10134, Torino -----------------Program (Speakers and abstracts)------------------ 14:30-15:10 Tiziano De Angelis (School of Mathematics, University of Leeds) Title: Probabilistic results concerning smoothness of the value function and of the free boundary in optimal stopping Abstract: I will present probabilistic proofs of some regularity properties for the value function of general optimal stopping problems and for the associated optimal boundaries. In particular this talk focuses on C^1 regularity of the value function and Lipschitz continuity of the optimal boundary. Most of our arguments rely on fundamental concepts from the theory of Markov processes and bridge the probabilistic and the analytical strands of the literature on free boundary problems. I will also illustrate situations in which our work improves or complements known facts from PDE theory. 15:10-15:40 Elena Issoglio (School of Mathematics, University of Leeds) Title: A numerical scheme for a multidimensional SDEs with distributional drift Abstract: This talk focuses on a multidimensional SDE where the drift is an element of a fractional Sobolev space with negative order, hence a distribution. This SDE admits a unique weak solution in a suitable sense - this was proven in [Flandoli, Issoglio, Russo (2017)]. The aim here is to construct a numerical scheme to approximate this solution. One of the key problems is that the drift cannot be evaluated pointwise, hence we approximate it with suitable functions using Haar wavelets, and then apply (an extended version of) Euler-Maruyama scheme. We then show that the algorithm converges in law, and in the special 1-dimensional case we also get a rate of convergence. This talk is based on a joint work with T. De Angelis and M. Germain. 15.40-16:10 coffee break 16:10-16:50 Elena Vigna (ESOMAS , University of Leeds) Title: On time consistency for mean-variance portfolio selection Abstract: This talk addresses a comparison between different approaches to time inconsistency for the mean-variance portfolio selection problem. We define a suitable intertemporal preferences-driven reward and use it to compare the three possible approaches to time inconsistency for the mean-variance portfolio selection problem over $[t_0,T]$: precommitment approach, consistent planning or game theoretical approach, and dynamically optimal approach. We find that the precommitment strategy beats the other strategies if the investor only cares at the view point at time $t_0$ and is not concerned to be time-inconsistent in $(t_0,T)$; the consistent planning strategy dominates the dynamically optimal strategy until a time point $t^*\in(t_0,T)$ and is dominated by the dynamically optimal strategy from $t^*$ onwards. 16:50-17:30 Cristina Zucca (Dept. of Mathematics, University of Torino) Title: Inverse First Passage Time problem for diffusions Abstract: In several applications the dynamics of the variables of interest is described via suitable stochastic processes constrained by boundaries and the focus is on the first passage time (FPT). This is the direct FPT problem. However, there are also instances when the underlying stochastic process is assigned, one knows or estimates the FPT distribution and wishes to determine the corresponding boundary shape. This is the inverse first passage time (IFPT) problem. Here we study this problem in the case of one dimensional diffusion process constrained by a single boundary or two boundaries. We also extend the IFPT method to multivariate Gauss-Markov processes and we investigate the boundary shape corresponding to given FPT distributions for suitable choices of the parameters. Special attention is given to the Integrated Brownian motion and the two-dimensional Ornstein-Uhlenbeck process. Some examples of applications of the proposed methods will be illustrated. The participation is free. Please, share this information with people potentially interested. -- %------------------------------------------------------- Elvira Di Nardo Dept. Mathematics "G. Peano" University of Torino Via Carlo Alberto 10 10123 Torino, Italia tel. +39 0116702862 fax +39 0116702878 http://www.elviradinardo.it %------------------------------------------------------- <http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_cam…> Mail priva di virus. www.avg.com <http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_cam…> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

1 0

Seminari 25 marzo a Pisa
by Dario Trevisan 23 Mar '19

23 Mar '19

Cari tutti, lunedì 25 marzo presso la Sala Seminari del Dipartimento di Matematica dell'Università degli studi di Pisa si terranno i seguenti seminari: (Dear all, on Monday, March 25th, the following two talks will take place in Sala Seminari, Dipartimento di Matematica, University of Pisa:) ORE 14:00 -- Stefano Pagliarani (Università degli studi di Udine) **Contraction methods for a class of McKean-Vlasov (mean-field) SDEs with jumps.** ORE 15:00 -- Martin Saal (University of Darmstadt) **White noise solutions for (m)SQG** Tutti gli interessati sono invitati a partecipare. ### Abstract S. Pagliarani: We consider two prototype classes of McKean-Vlasov (mean-field) SDEs with jumps. In the first case, the coefficient is assumed to be affine in the state-variable, only measurable in the law, and the dynamics allow for Lévy jumps. We study the equivalent functional fixed-point equation for the unknown time-dependent coefficients of the associated Markovian SDE. By proving a contraction property for the functional map in a suitable normed space, we infer existence and uniqueness results for the MK-V SDE, and derive a discretized Picard iteration method that approximates the law of the solution. Numerical illustrations show the effectiveness of the method, which appears to be appropriate to handle multi-dimensional settings. We finally describe possible extensions and generalizations to more general settings. The second class of MKV SDEs that we consider allows for self-exicitng jumps, which amounts to having jumps through hitting the boundary in the equivalent large particle system. The corresponding PDE problem is a particular instance of free-boundary problems known as supercooled Stefan-like problems. We write a Volterra equation for the free boundary and prove contraction results that are useful to determine the properties of the solution and to provide convergent numerical schemes. Abstract M. Saal: The inviscid surface quasigeostrophic equation (SQG) describes (roughly speaking) the temperature in a rapidly rotating stratified fluid which is transported by the velocity field. The velocity field is connected to the temperature via Riesz-transform, which are singular integral operators. It has applications in both meteorological and oceanic flows, while in mathematics it is often used as a toy model for the 3D Euler equations due to some structural similarities of these equations. We give a brief overview on versions of the SQG equation and of the known mathematical results. For a modified version (mSQG) with a smoother velocity field, which links the SQG equation to the vorticity formulation of the 2D Euler equations, we will show that by using a special symmetry in the kernel a white noise solution to mSQG can be constructed. Finally, we give some comments on the difficulties of our approach in the case of SQG itself. -- Dario Trevisan, Università degli Studi di Pisa, Dipartimento di Matematica, Largo Bruno Pontecorvo 5, 56127 - Pisa (PI) Italy telephone: (+39) 050 2213832 mobile: (+39) 331 2899761 Skype: dario-trevisan e-mail: dario.trevisan AT unipi.it webpage: http://people.dm.unipi.it/trevisan/

1 0

RTDA 13/D4 - Metodi matematici dell'economia e delle scienze attuariali e finanziarie
by Marco Scarsini 23 Mar '19

23 Mar '19

Bando Ricercatore art. 24, comma 3, lett. a) 13/D4 - Metodi matematici dell'economia e delle scienze attuariali e finanziarie Pubblicato sulla Gazzetta Ufficiale in data 15 marzo 2019 Settore scientifico-disciplinare SECS-S/06 - Metodi matematici dell'economia e delle scienze attuariali e finanziarie, presso il Dipartimento di Economia e Finanza della Luiss Libera Università Internazionale degli Studi Sociali Guido Carli. Codice concorso DEF-RIC-01/2019 Scadenza: 29 aprile 2019 - ore 14:00 Central European Summer Time (CEST) UTC+2 https://www.luiss.it/sites/www.luiss.it/files/Bando_7.pdf ******************************************************* Marco Scarsini Dipartimento di Economia e Finanza Luiss Viale Romania 32 00197 Roma, ITALY URL: http://docenti.luiss.it/scarsini/

1 0

corso di dottorato a roma1
by Faggionato, Alessandra 23 Mar '19

23 Mar '19

Buongiorno Lunedì 1 aprile, ore 10:30-13:00, aula B, Dipartimento di Matematica La Sapienza ci sara' il primo incontro del corso di dottorato di Fabio Toninelli (in cui verra' anche stabilito l'orario definitivo). Il corso si svolgerà durante il mese di aprile. sotto trovate dettagli del corso. saluti alessandra faggionato %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Fabio Toninelli (University Lyon 1) The dimer model: equilibrium and non-equilibrium aspects Abstract. This course focuses on various mathematical aspects of lattice dimer models. These are very classical two-dimensional statistical mechanics models, that are exactly solvable in some sense (Kasteleyn, 1961): partition function and correlations can be computed in determinantal form. Recently there has been a new wave of interest in dimer models, both in probability, combinatorics and mathematical physics. One reason is that these models, as well as other two-dimensional critical models, exhibit conformal invariance properties. Another interesting aspect is that they allow to obtain very nice Markov dynamics of two-dimensional interfaces, whose large-scale dynamical behavior can be studied. Detailed contents: • Kasteleyn theory (partition functions, correlations, determinantal properties) • correlations and representation determinantale • thermodynamic limit and ergodic Gibbs measures • height fluctuations and massless Gaussian field • dynamics of dimer models: mixing time and hydrodynamic limits. References : • R. Kenyon, Lectures on dimers, arXiv:0910.3129. • F. Toninelli, Lecture notes on the dimer model, http://math.univ-lyon1.fr/homes-www/toninelli/noteDimeri.pdf-- ************************************************* Prof. Alessandra Faggionato http://www1.mat.uniroma1.it/~faggionato/ Department of Mathematics University "La Sapienza" Piazzale Aldo Moro, 5 00185 - Rome Office 5, Phone (0039) 06 49913252 *************************************************

1 0

TWELFTH EUROPEAN SUMMER SCHOOL IN FINANCIAL MATHEMATICS
by Tiziano Vargiolu 22 Mar '19

22 Mar '19

(please disregard the other message, this one is correct) We are pleased to announce the TWELFTHTH EUROPEAN SUMMER SCHOOL IN FINANCIAL MATHEMATICS that which will be held at the University of Padova, September 2 - 6, 2019. The European Summer School in Financial Mathematics aims at bringing together the most talented young researchers in the field, looking at the very young who only just started their PhD studies. The Summer School is centered around the two courses: - Gilles Pages, Benedikt Wilbertz: "Numerical Probability" - Mark Podolskij, Roberto Reno': "High Frequency Data" For schedule, registration and application for financial support, see: https://www.math.unipd.it/~vargiolu/ESSFM/ On behalf of the organizing committee Tiziano Vargiolu -------------------------------------------------------------------------- Tiziano Vargiolu Dipartimento di Matematica Phone: +39 049 8271383 Universita' di Padova Fax: +39 049 8271428 Via Trieste, 63 E-mail: vargiolu(a)math.unipd.it I-35121 Padova (Italy) WWW: http://www.math.unipd.it/~vargiolu --------------------------------------------------------------------------

1 0

TWELFTH EUROPEAN SUMMER SCHOOL IN FINANCIAL MATHEMATICS
by Tiziano Vargiolu 22 Mar '19

22 Mar '19

We are pleased to announce the TWELFTHTH EUROPEAN SUMMER SCHOOL IN FINANCIAL MATHEMATICS that which will be held in Vienna, University of Padova, September 2 - 6, 2019. The European Summer School in Financial Mathematics aims at bringing together the most talented young researchers in the field, looking at the very young who only just started their PhD studies. The Summer School is centered around the two courses: - Gilles Pages, Benedikt Wilbertz: "Numerical Probability" - Mark Podolskij, Roberto Reno': "High Frequency Data" For schedule, registration and application for financial support, see: https://www.math.unipd.it/~vargiolu/ESSFM/ On behalf of the organizing committee Tiziano Vargiolu -------------------------------------------------------------------------- Tiziano Vargiolu Dipartimento di Matematica Phone: +39 049 8271383 Universita' di Padova Fax: +39 049 8271428 Via Trieste, 63 E-mail: vargiolu(a)math.unipd.it I-35121 Padova (Italy) WWW: http://www.math.unipd.it/~vargiolu --------------------------------------------------------------------------

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

Random