- Random - Mailing list — Dipartimento di Matematica, UniPI

Avviso di mini-corso :: C.V.D.@DSS (Scienze Statistiche, Sapienza)
by Pierpaolo Brutti 31 Mar '14

31 Mar '14

----------------------------------------------------------------------------- A v v i s o d i M i n i - C o r s o ----------------------------------------------------------------------------- Venerdì 04 Aprile, ore 10:00-13:00 ----------------------------------------------------------------------------- Aula VI (quarto piano) Dipartimento di Scienze Statistiche Sapienza Università di Roma C.LAUDIO DURASTANTI V.ALENTINA CAMMAROTA D.OMENICO MARINUCCI (Dip. di Matematica, Università degli Studi di Roma Tor Vergata) terranno un mini-corso dal titolo STEIN-MALLIAVIN APPROXIMATIONS IN STATISTICS & PROBABILITY tutti gli interessati sono invitati a partecipare. ----------------------------------------------------------------------------- Maggiori informazioni sui seminari presso il DSS sono consultabili a quest'indirizzo: http://goo.gl/Y6OQYm Saluti Pierpaolo Brutti - Fulvio De Santis --- P R O G R A M Part I: The Geometry of Needlets Excursion Sets Domenico Marinucci In this talk, we shall be concerned with geometric functionals and excursion probabilities for some nonlinear transforms evaluated on wavelet/needlet components of spherical random fields. For such fields, we consider smoothed polynomial transforms, such as those arising from local estimates of angular power spectra and bispectra; we focus on the geometry of their excursion sets, and we study their asymptotic behaviour, in the high-frequency sense. Our main technical tools are recent results on Malliavin calculus and Total Variation bounds for Gaussian subordinated fields by Nourdin and Peccati, and the Gaussian Kinematic Formula by Adler and Taylor, which we shall review briefly in the talk. We put particular emphasis on the analysis of Euler-Poincaré characteristics, which can be exploited to derive extremely accurate estimates for excursion probabilities. The present analysis is motivated by the statistical investigation of asymmetries and anisotropies in Cosmic Microwave Background radiation (CMB) data. Based on joint work with Sreekar Vadlamani Part II: Stein-Malliavin Approximations for Needlets Polyspectra Valentina Cammarota We provide quantitative Central Limit Theorems, in the high frequency limit, for nonlinear transforms of spherical wavelets/needlets random fields, which are of immediate interest for spherical data analysis. In particular, we focus on so-called needlets polyspectra, which are popular tools for nonGaussianity analysis in the cosmological community, and on the area of excursion sets. Our results are based on Stein-Maliavin approximations for nonlinear transforms of Gaussian fields, and on an explicit derivation on the high-frequency limit of the fields' variances, which may have some independent interest. Based on joint work with Domenico Marinucci Part III: Normal Approximations for Linear and U-Statistics on Spherical Poisson Fields Claudio Durastanti We review a recent stream of research on Normal approximations for linear and U-statistics of wavelets/needlets coefficients evaluated on a homogeneous spherical Poisson fi...eld. We show how exploting results from Peccati and Zheng (2011), based on Malliavin calculus and Stein''s methods, it is possible to assess the rate of convergence to Gaussianity for a triangular array of statistics with growing dimensions. These results can be exploited in a number of statistical applications, such as spherical density estimations, searching for point sources, estimation of variance and the spherical two-sample problem. Based on joint work with Solesne Bourguin, Domenico Marinucci and Giovanni Peccati.

1 0

A Workshop to Celebrate the Statistical Contributions of A. Philip Dawid: Causal Inference, Graphical Models and Prediction, University of Cambridge, UK (30 April 2014)
by Julia Mortera 30 Mar '14

30 Mar '14

Dear Colleague, Many thanks to those who have already registered and who have emailed us their unavailability for the workshop. *** Gentle reminder please *** *Workshop:* The Statistical Contributions of A. Philip Dawid: Causal Inference, Graphical Models and Prediction *Venue:*** Centre for Mathematical Sciences, University of Cambridge, UK *Date:* 30th April 2014 *Dinner Date:* 29 April 2014 *Registration deadline:* 15 April 2014 * Registration and ***more information*: * https://sites.google.com/site/workshopdawid2014/ *Dinner only option:* If you wish to join us only for dinner, please let us know by emailing Hui Guo at hg283(a)cam.ac.uk by 15 April. You will be able to make payments for the dinner in the restaurant on the day. *Confirmed speakers:* Carlo Berzuini (University of Manchester) Simon Byrne (UCL) Panayiota Constantinou (University of Bristol) Vanessa Didelez (University of Bristol) Sara Geneletti (LSE) Peter Grünwald (CWI and Leiden University) Hui Guo* *(University of Cambridge) Steffen Lauritzen (University of Oxford) Monica Musio(University of Cagliari) Julia Mortera (Umiversity of Rome) David Spiegelhalter (University of Cambridge) Volodya Vovk (Royal Holloway University of London) Thanks for Your attention. We would be very happy if you could attend the meeting. Kind Regards, Luisa Bernardinelli Carlo Berzuini Panayiota Constantinou Vanessa Didelez Hui Guo Nuala Sheehan -- Prof.ssa Julia Mortera Dipartimento di Economia Università Roma Tre Via Silvio D'Amico 77 00145 Roma Tel: +39 06 57335732 Fax: +39 06 57335771 http://dipeco.uniroma3.it/docenti.asp?id=123# --- Questa e-mail è priva di virus e malware perché è attiva la protezione avast! Antivirus. http://www.avast.com

1 0

Seminario di Guido Montorsi (Roma - 11 aprile)
by Marco Isopi 29 Mar '14

29 Mar '14

Seminari MoMA www.mat.uniroma1.it/ricerca/seminari/moma Venerdi’ 11 aprile, ore 12-14, aula Consiglio Guido Montorsi Analog Digital Belief Propagation and its application to channel decoders Sunto As required by information theory, channel codes with long code words are necessary building blocks in a transmission system to achieve reliable communications with minimal power and maximal throughputs over noisy physical channels. Nowadays capacity-achieving large random binary codes are actually adopted in most telecommunication standards. The breakthrough that allowed their practical use has been to substitute the optimal maximum likelihood decoding techniques at the receiver with suboptimal iterative techniques based on “belief” propagation. Belief propagation is a powerful inference technique working on graphs used in many different applications. Graph nodes are associated to factors or constraints of the model while graph edges are associated to random variables. Belief propagation algorithm proceeds by iteratively updating messages associated to the random variables, according to the constraints imposed by nodes. In the framework of channel decoding the graph nodes represent the deterministic, usually linear, code constraints. They are either associated to a (binary) parity check sum, or to a repetition of the variable. Belief propagation is initialized with a set of messages obtained from the noisy channel observations of the transmitted bits and proceeds iteratively, updating the messages according to the code constraints until convergence is reached. Most practically used codes are linear and binary codes, so that messages propagated in the graph are binary messages usually represented with a single scalar (the Log-Likelihood Ratio). In order to increase the throughput of communication systems, the use of non binary codes is an attractive solution as each symbol can carry more information bits. Non binary codes can be constructed over groups, rings, or fields and there is a vast literature on the design of capacity achieving non binary codes. The extension of the application of belief propagation to the non binary codes however poses several complexity problems as both message representation and message updating at check node grows at least linearly with the cardinality of the non binary alphabet and consequently exponentially with the increase of required throughput. In this talk I will start by recalling the fundamental ideas and terminology beyond binary channel coding constructions and corresponding iterative decoding with belief propagation and other iterative techniques. I will then extend the concepts to non binary codes and summarize the main algorithms and complexity problems related with them. I will then introduce a new algorithm, named Analog Digital Belief Propagation (ADBP) which solves the complexity problems of belief propagation for non binary codes. I will discuss the main properties of the algorithm, its possible extensions and code design problems related to the adoption of this algorithm.

1 0

Seminario Lucian Maticiuc [ Dipartimento Informatica UniVr ]
by Luca di persio 28 Mar '14

28 Mar '14

Il giorno Mercoledì 9 Aprile, alle ore 17 [ rinfresco ore 16:45 ] presso il Dipartimento di Informatica Università degli studi di Verona , strada le Grazie 15, Verona Lucian Maticiuc "G. Asachi" Technical University, Romania terrà un seminario dal titolo Stochastic Differential Delay Systems and Optimal Control Problems *Abstract* We show the existence and the uniqueness of the solution for some multivalued stochastic differential equation with delay (the multivalued term is of subdifferential type). Afterwards we consider optimal control problems where the state X is a solution of a controlled delay stochastic system as above. We establish the dynamic programming principle for the value function and finally we prove that the value function is a viscosity solution for a suitable Hamilton- Jacobi-Bellman type equation. __ Luca Di Persio - PhD assistant professor of Probability and Mathematical Finance Dept. Informatics University of Verona strada le Grazie 15 - 37134 Verona - Italy Tel : +39 045 802 7968 Dept. Math University of Trento V. Sommarive, 14 - 38123 Povo - Italy Tel : +39 0461 281686

1 0

seminario di Daniele Marazzina presso Prometeia - Bologna
by Giacomo Scandolo 28 Mar '14

28 Mar '14

Il giorno Giovedì 3 Aprile 2014, alle ore 14:30 presso la sede di Prometeia (sala grande, primo piano) via G.Marconi 43, Bologna Daniele MARAZZINA (Politecnico di Milano) terrà un seminario dal titolo "Fast pricing of discretely monitored exotic options by a numerical Wiener-Hopf procedure" Abstract We present a fast and accurate pricing techniques based on the Spitzer identity and the Wiener-Hopf factorization. We apply it to barrier and lookback options when the monitoring is discrete and the underlying evolves according to an exponential Lévy process. The numerical implementation exploits the fast Fourier transform and the Euler summation. The computational cost is independent of the number of monitoring dates; the error decays exponentially with the number of grid points.

1 0

Seminario Lucian Maticiuc [ UniTn ]
by Luca di persio 28 Mar '14

28 Mar '14

*Giovedì 3 aprile **ore 10.00* *------------------------------ * *Aula Seminari - Dipartimento di Matematica* *------------------------------ * *"Backward S**tochastic Variational Inequalities with* *Time Delayed Generators"* *Lucian Maticiuc* *(**"Gheorghe Asachi" Technical University and "Alexandru Ioan Cuza" University, Iasi, Romania)* *ABSTRACT*: We show the existence and the uniqueness of the solution for the following multivalued backward stochastic differential equation withtime-delayed generators: Where is the subdifferential operator of a proper convex lower semicontinuous function *Referente*: *Luca Di Persio* -- __ Luca Di Persio - PhD assistant professor of Probability and Mathematical Finance Dept. Informatics University of Verona strada le Grazie 15 - 37134 Verona - Italy Tel : +39 045 802 7968 Dept. Math University of Trento V. Sommarive, 14 - 38123 Povo - Italy Tel : +39 0461 281686

1 0

Seminario ROUSSEAU
by Matteo Ruggiero 28 Mar '14

28 Mar '14

**************************************************************** STATISTICS SEMINARS @ COLLEGIO CARLO ALBERTO **************************************************************** Venerdi' 4 Aprile 2014 alle ore 12:00, presso la sala rossa del Collegio Carlo Alberto, Moncalieri (TO), si terra' il seguente seminario: Judith ROUSSEAU (Université Paris Dauphine) BEHAVIOUR OF THE POSTERIOR DISTRIBUTION IN HMM MODELS WHEN THE NUMBER OF STATES IS MISSPECIFIED Abstract: In this paper we study the asymptotic behaviour of the posterior distribution for parametric HMM models with finite number of components. We concentrate in particular on the case where the number of states of the hidden Markov chain in the model is larger than the true number of states. In such cases the parameter is not identifiable. We study the impact of the prior on the transition matrix on the asymptotic behaviour of the posterior distribution. In particular that if some independent Dirichlet priors are considered on each row of the transition matrix we can identify different types of behaviour of the asymptotic posterior distribution depending on the choice of the parameter of the Dirichlet distributions. A link is made with static mixture models and some numerical study is conducted to illustrate the results and investigate the behaviour of the posterior in cases where no theoretical results are provided. Tutti gli interessati sono invitati a partecipare. Il seminario e' organizzato dalla "de Castro" Statistics Initiative (http://www.carloalberto.org/stats) in collaborazione con il Collegio Carlo Alberto. Cordiali saluti, Matteo Ruggiero --------- Matteo Ruggiero University of Torino & Collegio Carlo Alberto http://web.econ.unito.it/ruggiero

1 0

seminario di Pablo Ferrari a Roma Tre
by pietro caputo 28 Mar '14

28 Mar '14

Martedi 1 Aprile, alle ore 14:30 in aula 311 del Dipartimento di Matematica e Fisica dell'Universita' di Roma Tre, il Prof. Pablo Ferrari (Buenos Aires) terra' un seminario su: "Yaglom limits via Holley inequality" Abstract: We consider Markov chains on a countable partially ordered state space with an absorbing state. Assume that the absorbed chain has a quasi stationary distribution. We give sufficient conditions on the transition probabilities to guarantee that the law of the chain at time n conditioned to non absorption is monotone non decreasing in n. As a consequence the Yaglom limit (the limit of the conditioned chain starting from a minimal state, as n goes to infinity) converges to a quasi stationary distribution; the limit is minimal for the stochastic order of measues. The approach uses a dynamics on the space of trajectories and Holley inequality. Joint work with Leonardo Rolla.

1 0

Mercati Energetici e Metodi Quantitativi - edizione 2014
by Tiziano Vargiolu 25 Mar '14

25 Mar '14

(chiedendo scusa in anticipo per eventuali ripetizioni) Buongiorno, volevamo ricordare che la scadenza per presentare contributi al workshop "Mercati Energetici e Metodi Quantitativi: un Ponte tra Universit? e Aziende" e' MARTEDI' 1. APRILE. Tutte le informazioni sul sito web http://www.unponte2014.stat.unipd.it/ dove si possono anche trovare gli interventi gia' accettati: http://www.unponte2014.stat.unipd.it/programma.htm Inoltre ricordiamo che quest'anno abbiamo deciso di estendere l'iniziativa organizzando, nei due giorni precedenti la conferenza, quattro tutorial sui seguenti argomenti: - Introduzione ai mercati dell'energia e del gas - Rischio di credito e di controparte nei mercati energetici - Strumenti di analisi e di previsione per i mercati energetici. - Tecniche di gestione di portafoglio e di ottimizzazione: i contratti strutturati nei mercati dell'energia. Tutte le informazioni sui tutorial si trovano alla pagina http://events.math.unipd.it/memq2014/ In attesa di rincontrarci a Padova, porgiamo i nostri pi? Cordiali Saluti, gli Organizzatori Giorgia Callegaro Enrico Edoli Francesco Lisi Tiziano Vargiolu

1 0

Avviso di seminario :: Liseo@DSS (Scienze Statistiche, Sapienza)
by Pierpaolo Brutti 25 Mar '14

25 Mar '14

----------------------------------------------------------------------------- A v v i s o d i S e m i n a r i o ----------------------------------------------------------------------------- Venerdì 28 Marzo, ore 11am ----------------------------------------------------------------------------- Stanza 34 Dipartimento di Scienze Statistiche Sapienza Università di Roma BRUNERO LISEO (Dip. di metodi e modelli per il territorio, l'economia e la finanza, Sapienza Università di Roma) terrà un seminario dal titolo APPROXIMATE BAYESIAN INFERENCE FOR COPULA MODELS tutti gli interessati sono invitati a partecipare. ----------------------------------------------------------------------------- Maggiori informazioni sui seminari presso il DSS sono consultabili a quest'indirizzo: http://goo.gl/Y6OQYm Saluti Pierpaolo Brutti - Fulvio De Santis --- Summary Si propone un metodo per la stima di un funzionale di interesse relativo ad un modello multivariato. Il metodo è basato su una rappresentazione in termini di copula e sfrutta alcuni recenti sviluppi delle tecniche computazionali bayesiane approssimate che utilizzano metodi di verosimiglianza empririca. Il termine "approximate" nel titolo è dovuto a due fattori: i) il processo di elicitazione è limitato al solo funzionale di interesse e alle distribuzioni marginali. ii) Non si utilizza la "vera" funzione di verosimiglianza, bensì una sua approssimazione empirica (Owen, 2010).

1 0