Forthcoming Meetings
Here is a list of scientific activities which are being organized or co-organized by members of our group.
Mini course: Generalized Gibbs measures in Mathematical statistical mechanics, by Arnaud Le Ny - Université de Paris-Est
Organizer: Sandro Gallo
Dates: 16 Nov (14h00 - 16h00 • Room 4-005 ICMC-USP) and 17 Nov 2016 (14h00 - 16h00 • Room 4-111 Auditório Fávaro ICMC-USP) São Carlos, SP
Abstract: In this talks, we shall introduce the DLR description of measures on infinite product probability space and explain how Gibbs measures that arises naturally as equilibrium states in this context can provide a natural description of phase transitions at infinite volume. Using basic examples of Ising models in dimension one, two or three, or on Cayley trees, we shall review the main features of this theory and study the (un-)stability of the Gibbs property under scaling transformations of the so-called renormalisation group. Putting this unstability - or failure of Gibbs property - in a nutshell via the fundamental example of the decimations of Ising models, we shall describe also the main features of the Dobrushin program of restoration of Gibbsianness and provide in dimension one a few hints to the relationsips between g-measures and proper DLR Gibbs measures.
A day on statistics and applied probability
Organizers: Carlos Diniz, Sandro Gallo, Rafael Stern, Pablo Rodriguez
Date: Nov 18 2016 • Auditório 3 next to the Biblioteca Comunitária da UFSCar, SP
Site: http://scpc.icmc.usp.br/dosap/
Organized by the Argentinean Association for Industrial, Computational and Applied Mathematics.
Date: May 02 - 05 2017 • Comodoro Rivadavia, Chubut, Argentina
Past Activities
The following activities have been organized or co-organized by members of our group (since 2014).
Mini course: Variable Length Markov Chains and g-Measures, by Frédéric Paccaut - Université de Picardie Jules Verne
Organizer: Sandro Gallo
Dates: 09 and 10 Nov 2016 (14h00 - 16h00 • Room 5-104 ICMC-USP) and 11 Nov 2016 (16h00 - 18h00 • 4-111 Auditório Fávaro ICMC-USP) São Carlos, SP
Abstract: This minicourse will deal with two families of similar random processes: g-measures and variable length markov chains (VLMC). Two different points of view will be used to study these processes. We will be mainly interested in the problem of existence of such processes, in full generality.
- For g-measures, we will adopt the formalism of ergodic theory (Perron-Frobenius operators, variational principal). We will give results for discontinuous transitions and brie y explain the connections with dynamical systems.
- For the variable length Markov chains, we will take advantage of the combinatorial structure of the context tree used to define the chain.
- Finally, as an application, we will use simple VLMC as sources to produce words stored in a data structure called sux trie. We will discuss the typical height and depth of these tries.
Mini course: An Introduction to Fractional Calculus and its Relationship to Stochastic Processes, by Federico Polito - Università di Torino
Organizer: Pablo Rodriguez
Date: 22 Aug 2016 • Time: 10h00 - 11h40, 14h00 - 15h30, 15h50 - 17h20 • 4-111 Auditório Fávaro ICMC-USP São Carlos, SP
Abstract: In the first part of the course, after a brief review of the basic tools of mathematical analysis and probability we are going to use, I will introduce the basic definitions and properties of fractional calculus operators commonly used in applied mathematics. Some generalizations of the exponential function linked to them will be analyzed in detail. In the second part I will relate those operators to some stochastic processes of the Lévy class. Finally, some examples of applications will be discussed.
Pre-requisites: Mathematical Analysis, Probability, Basic Stochastic Processes.
Special Session: Stochastic processes in random environment and applications
In the 1st Joint Meeting Brazil-Italy in Mathematics. Organizers: Fabio Machado, Pablo Rodriguez and Fabio Zucca.
Dates of the session: Aug 29 - 30 2016 • IMPA Rio de Janeiro, RJ
Abstract: Random structures typically consist of many mutually interacting random components. Even if the interaction is local and simple, the collective behavior of a large number of degrees of freedom may show intriguing phenomena like anomalous fluctuations, phase transitions, complex time evolutions and metastability. Systems of this type play an important role in statistical physics, biology and theoretical computer science. Moreover, it is widely recognized that a deep mathematical understanding of these systems is crucial even for concrete applications. The aim of this session is to present some classic results and new challenges on this topics both from the theoretical and applied points of view. We expect to address a potential audience ranging from experts of the area to PhD students. The main topics are: interacting particle systems, Markov processes, disordered systems, statistical physics, stochastic models in biology and computer science.
Speakers:
- Christophe Gallesco (UNICAMP)
- Daniela Bertacchi (Università degli Studi di Milano-Bicocca)
- Enrique Andjel (IMPA/)
- Federico Polito (Università degli Studi di Torino)
- Heinrich Matzinger (Georgia Institute of Technology)
- Ido Ben-Ari (University of Connecticut)
- Serguei Popov (UNICAMP)
XX Escola Brasileira de Probabilidade
Organizers: Daniel Ahlberg, Paulo Faria da Veiga, Sandro Gallo, Roberto Imbuzeiro, Leandro Pimentel, Renato Fontes
Date: Jul 04 - 09 2016 • ICMC-USP São Carlos, SP
Site: www.impa.br/opencms/en/eventos/store_2016/evento_1613
2o Dia temático. NeuroMat: Métodos estatísticos, probabilísticos e computacionais em Neurobiologia.
Organizers: Sandro Gallo, Rafael Stern, Antonio Galves
Date: 03 Jun 2016 • Time: 10h00 - 16h00 • Auditório 3 Anexo da Biblioteca Comunitária da UFSCar, São Carlos, SP
Talks:
- Apresentação de um novo modelo estocástico para o sistema neuronal. Jorge Stolfi (UNICAMP)
- Estudos de simulação de um novo modelo estocástico para o sistema neuronal. Antonio C. Roque (USP Ribeirão Preto)
- O cérebro estatístico e o jogo do goleiro. Bruno M. de Castro (doutorando NeuroMat/USP SP) e Marcos D. Gubitoso (USP SP)
- Questões probabilísticas e estatísticas sugeridas pela neurobiologia. Antonio Galves (USP)
Programa
Stochastic chains: hitting, returning and long-range dependence
Organizer: Sandro Gallo
Date: 20 Nov 2015 • Time: 09h00 - 17h20 • DEs-UFSCar São Carlos, SP
Talks:
- Hitting times in a leakage and fire model. Miguel Abadi (IME-USP)
- A martingale approach to waiting time problems in i.i.d. and Markov trials. Renato Gava (UFSCar)
- On the shortest path between two observables. Rodrigo Lambert (UFU)
- First Return Time of the sequence under β - Mixing conditions. Alejandra Rada (IME-USP)
- Dynamic uniqueness and phase transition of chains of infinite order. Daniel Takahashi (Princeton)
- Projective Convergence of measures. Edgardo Ugalde (San Luis Potosí)
Concentration Inequalities in Probability: a crash course, by Jean-René Chazottes - École Polytechnique Paris
Organizer: Sandro Gallo
Dates: 16 - 19 Nov 2015 • Time: 16h00 - 18h00 • ICMC-USP São Carlos, SP
CIMPA School New interactions of Combinatorics and Probability
Organizers: Paulo da Veiga, Frédéric Patras, Igor Mencattini, and Pablo Rodriguez
Dates: Aug 24 - Sep 04 2015 • ICMC-USP São Carlos, SP
Site: cimpa-nicp.icmc.usp.br
XIX Brazilian School of Probability
Organizers: Cristian Coletti, Renato Gava, Fabio Machado, Serguei Popov, and Pablo Rodriguez
Dates: Aug 03 - 08 2015 • Maresias, SP
Site: www.ime.usp.br/ebp/ebp19
Curso: Métodos estatísticos para a análise de conectividade neural, por Daniel Yasumasa Takahashi - Princeton University
Organizadores: Sandro Gallo e Pablo Rodriguez
Dates: 11 - 13 Nov 2014 • Time: 09h00 - 12h15 • ICMC-USP São Carlos, SP
Resumo: O comportamento humano éé uma propriedade que emerge da interação de várias áreas do cérebro, partes do corpo, e o ambiente. Compreender como esses elementos interagem de forma dinâmica é um dos principais temas da neurociência de sistemas. Vários métodos de análise de séries temporais multivariadas foram introduzidas para descrever a interdependência entre áreas do cérebro usando sinais biológicos como EEG, BOLD, MEG, e LFP, e sõo chamados coletivamente de medidas de conectividade. O objetivo deste curso é introduzir o aluno a alguns métodos de séries temporais multivariadas projetados para responder a algumas perguntas neurocientíficas.
4th Workshop in Stochastic Modeling
Organizers: Cristian Coletti, Sandro Gallo, Renato Gava, Fabio Machado, Karina de Oliveira, Pablo Rodriguez, Paulo da Veiga, and Cristel Vera
Dates: 05 - 07 Nov 2014 • ICMC-USP São Carlos, SP
Site: wsm.icmc.usp.br
Thematic Session: Probability and Stochastic Processes
In 21 SINAPE. Organizers: Fabio Machado and Pablo Rodriguez
Dates: 20 - 25 Jul 2014 • Natal, RNP
Speakers:
- Sandro Gallo (UFSCar)
- Rinaldo Schinazi (University of Colorado)
- Daniel Valesin (University of British Columbia)
Course: Random directed trees and their limits, by Rahul Roy - Indian Statistical Institute, Delhi
Organizer: Pablo Rodriguez
Dates: 3, 4, 11, 12 Jun 2014 • Time: 08h00 - 10h00 • ICMC-USP São Carlos, SP
Abstract: In recent times mathematicians have started studying the minimal directed spanning trees (MDST) and the directed spanning forests (DSF). These models were introduced by geologists to study river basin geomorphology and later taken up by physicists in their study of online nearest neighbour graphs as examples of models which exhibit power law behaviour. Here two distinct kind of problems are studied. The first one concerns the MDST on the unit cube and the second one is the drainage model on the d-dimensional Euclidean space. We discuss the results known in these models.
Course: Information Theory, Probability and Statistics, by Yuri Suhov - University of Cambridge / ICMC-USP
Organizers: Paulo da Veiga and Pablo Rodriguez
Dates: 19 May - 16 Jun 2014 • ICMC-USP São Carlos, SP