TABLE OF CONTENTS - CIMAT

TABLE OFCONTENTS

5 Welcome Message

6 Administrative Program

7 Social Program

9 General Information

13 Program

Welcome MessageOn behalf of the Organizing Committee, I welcome you to Mérida and the 15th Latin American Congress on Probability and Mathematical Statistics. We are truly honored by your presence and sincerely hope that the experience will prove to be professionally rewarding, as well as memorable on a personal level. It goes without saying that your participation is highly significant for promoting the development of this subject matter in our region.

Please do not hesitate to ask any member of the staff for assistance during your stay, information regarding academic activities and conference venues, or any other questions you may have. The city of Mérida is renowned within Mexico for its warm hospitality, amid gastronomical, archaeological, and cultural treasures. Do take advantage of any free time to explore the city and its surroundings!

Dr. Daniel HernándezChair of the Organizing Committee, XV CLAPEM

6 c l a p e m 2 0 1 9 . e v e n t o s . c i m a t . m x

REGISTRATIONThe registration will be at the venue Gamma Mérida El Castellano Hotel. There will be a registration desk at the Lobby.

Sunday, December 1, 2019From 16:00 hrs to 22:00 hrs.

Monday, December 2, 2019From 8:00 to 17:00 hrs.

Tuesday, December 3, 2019From 8:00 to 17:00 hrs

Thursday, December 4, 2019From 8:00 to 16:00 hrs.

D O Y O U N E E D A N I N V O I C E F O R Y O U R R E G I S T R AT I O N F E E ?

Please ask for it at the Registration Desk at your arrival.

COFFEE BREAKS

During the XV CLAPEM there will be coffee break services which will be announced in the program and will be displayed in both venues: Gamma Mérida El Castellano Hotel and CCU, UADY.

Administrative Program


1) Opening ceremony & welcoming cocktail

Monday, December 2, 2019Auditorio “Manuel Cepeda Peraza”Centro Cultural Universitario (CCU), UADY17:00 hrs.“Vaquería Yucateca”Performance by Ballet Folclórico – Universidad Autónoma de Yucatán

2) Conference dinner

Thursday, December 5, 2019Started: 21:00 hrs.

Hacienda YA-AXKAHacienda Ya-axka, named after it characterised blue colour. A peculiar colour of the ancient Mayan cities which is represented on the mural paintings in the archaeological sites of the region. Inspired by the majesty and fascination of the haciendas of our beautiful state of Yucatan, recreating the old main houses of the henequen haciendas which reached their maximum splendour at the endo of the 20th century.

The dinner is a three course meal that includes two glasses of wine, two regionals beers, tequila, live music and round trip transportation to the city center of Merida.

* The cost of the dinner is $50 USD and it can be paid at the registration desk from Sunday 1st of December up to 14:00 hrs of Wednesday 4th of December.

Transportation to conference dinner

Please, show your conference dinner ticket when you get on the bus.

Downtown – Hacienda Ya-axka - Downtown

Thursday, December 5, 2019Downtown to Hacienda Ya-askaDeparture time: 20:15 hrs.Departure site: Parking cross the street from Hotel Gamma Mérida El CastellanoHacienda Ya-aska to DowntownBus 1: departure time: 22:30 hrs.Bus 2: departure time 24:00 hrs.Bus 3 and Bus 4: departure time: 02:00 hrs.

Social Program


General Information


General Information

Restaurant Address Food COST (MXP)

El Trapiche 62 St # 491Downtown Traditional food$

Cafetería Pop 57 St #501 between 60 St and 62 St Mexican food $

La Chaya Maya 62 St. # 481 x 57 Downtown Traditional food $$

100% Natural 62 St # 498 X 59 and 61Downtown Vegetarian food

$$

Los trompos 59 St # 502 X 60Downtown Mexican tacos

$$

Amaro 59 St # 507 X 60 and 62Downtown Italian food

$$

Pancho´s 59 St #509Santa Lucía Park Mexican food

$$

El Marlín Azul 62 St #488BSanta Lucía Park Sea food

$$

Bristrola 57 60 St # 488 X 57(Downtown) Restaurant & bar

$$

Peruano 55 St # 502 X 60 St and 62 St(Downtown) Peruvian food

$$

Pita Mediterranean Cuisine & Bar

55 St # 496Santa Lucía Park Mediterranean food $$

La Tratto Santa Lucía 60 St # 471 X 53 St and 55 St Italian food $$

Don Spaghetto Centro

Departamento a X 63 St and 65 StDowntown Italian food $$

Piñuela 60 St #491 between 55 St and 57 St(Downtown) Alta cocina $$$

La Pigua Av Cupules 62 Sea food $$$

WHERE CAN YOU EAT?

BANKS (near venues)


General Information

WHAT TO DO IN MERIDA CITY DURING EVENINGSMondayVaquería Yucateca (Tradicional regional dance of Yucatán)City House

21:00 hrs.

Tuesday “Remembranzas Musicales” (Popular music from the 1940 mostly Danzon)Santiago’s park, Downtown

20:30 hrs.

WednesdayRecorrido por el Cementerio de Mérida (Guide tour to the cementery of Merida)20:00 hrs.

“Noche de las Culturas” meeting Francisco de Montejo (Cultural Nights)Montejo’s house (Downtown)

20:30 hrs.

ThursdaySerenata (Serenade)Santa Lucía park

21:00 hrs.“En el Corazón de Mérida”From Thursday to Saturday

From 20:00 to 2:00 hrsIt consists of closing a complete street of the city center so that restaurants, musical groups, local artists and artisans can exhibit their products in full public roads during a period of time of three hours.

FridayVideo mapping “Piedras Sagradas”Cathedral of Merida

20:30 hrs.

SaturdayMexican night At the end of Paseo Montejo

Starting at 20:00 hrs.

Mayan ball game “Pok Ta Pok”

20:30 hrsBesides the cathedral of Merida

We hope you will enjoy this CLAPEM and Mérida City!!!

Emergency Phone: 911

WEATHER 30°C / 18°C (86°F/64°F)

At this time of the year, temperature at Mérida is about

Program


Monday, December 2

PROGRAM

Auditorium, CCU-UADY

8:30 – 9:30 Plenary Talk

Average Gromov hyperbolicity and the Parisi ansatz, Sourav Chatterjee (Stanford University, US)

9:30 – 12:00Thematic Session

TS Random processes on networks and their limitsOrganizer: Louigi Addario-Berry (MacGill University, CA) Christina Goldschmidt, (University of Oxford, UK)The scaling limit of a critical random directed graph

Anja Sturm (Georg-August-Universität Göttingen, DE)Recursive tree processes and mean-field limits of interacting particle systems

(10:30 – 11:00 Coffee break - both venues)

Roberto Imbuzeiro Oliveira, (IMPA, BR) Interacting diffusions over random graphs and the role of sparsity

Simon Griffiths (PUC-Río, BR)Moderate deviation probabilities for subgraphs and other discrete structures

Hall 1, CCU-UADY

9:30 – 12:00Contributed Session

CS, Random dynamical systems with jumps I Organizer: Juan Carlos Pardo Millán (CIMAT, MX)

Random events occur in nature throughout our everyday experiences. Taking stochastic effects into account is of central importance for the development of mathematical models of complex phenomena under uncertainty arising in applications. Macroscopic models in the form of differential equations for these systems contain randomness in many ways,


Monday, December 2

such as stochastic forcing, uncertain parameters, random sources or inputs, and random initial and boundary conditions. The theory of random dynamical systems and stochastic differential equations provides fundamental ideas and tools for the modeling, analysis, and prediction of complex phenomena. The aim of this contributed session is to showcast new developments on the theory of random dynamical systems whose driven noise has a jump structure. The three speakers on this session will deal with different aspects of stochastic differential equations driven by Lévy processes. Gerardo Barrera, (University of Alberta, CA) Cut-off phenomenon for Ornstein--Uhlenbeck processes driven by Lévy processes Joaquín Fontbona (Universidad de Chile, CL)Synchronization of stochastic mean field networks of Hodgkin-Huxley neurons with noisy channels

Ilya Pavlyukevich, (University of Jena, DE)Lévy-driven transport equations

Hall 2, CCU-UADY

9:30 – 12:00 Contributed Talks

CT, Felipe Muñoz (Universidad de Chile and Université Paris-Saclay, CL)Rate of convergence of a spatial branching process in the large population limit using optimal transport

CT, Laura Eslava, (IIMAS, UNAM, MX)Branching processes with cousin merges and locality of hypercube´s critical percolation


CT, Freddy Palma Mancilla (IMUNAM, MX)Intertwining of Galton-Watson processes

CT, Peter Pflaumer (TU Dortmund, DE)Euler´s and Süßmilchs´s Population Growth Model

Hall 3, CCU-UADY


CT, Alexander Alvarez (University of Prince Edward Island, CA)A robust approach to construct coherent risk measures

CT, Laszlo Markus (Eötvös Loránd University, Budapest, HU)Stochastic correlation for modelling association of stock prices



CT, Jonathan Chavez Casillas (University of Rhode Island, US)A level-1 Limit Order book with time dependent arrival rates

CT, Adriana Ocejo (University of North Carolina at Charlotte, US)Optimal investment portfolio of a variable annuity policyholder

Hall 1, Gamma Mérida El Castellano Hotel

9:30 – 12:00 Contributed Session

CS, Probabilistic Methods in PDEs and SPEDEs Organizer: José Alfredo López Mimbela (CIMAT, MX)

Asymptotic properties of semi-linear partial differential equations, such as explosion in finite time an existence of solutions defined globally in time, is a current and very active field of research in Mathematics. Moreover, existence and regularity results of solutions of stochastic partial differential equations perturbed by Gaussian noises more general than Brownian motion has been intensively studied in recent years. Probabilistic methods have been shown to be a powerful tool to understand these properties, both in PDEs and SPDEs, including reaction difussion equations with non-autonomous generators, or generators of anomalous difussions and systems of semi-linear PDEs and SPDEs. The aim of this session is to report several recent developments in these fields, where the probabilistic methods are crucial.

Aroldo Pérez Pérez, (Universidad Juárez Autónoma de Tabasco, MX)Blow up of fractional reaction-diffusion systems with and without convection terms

Ekaterina Todorova Kolkovska (CIMAT, Guanajuato, MX)Survival of semilinear equations related to certain measure-valued Markov branching processes”


Eugenio Guerrero Ruiz, (Universidad del Caribe, MX)Pointwise eigenfunction estimates and mean Lp-norm blowup of a system of semilinear SPDEs with symmetric Lévy generators

José Alfredo López-Mimbela (CIMAT, Guanajuato, MX)Large time behavior of semilinear SPDEs driven by a Brownian Motion

Monday, December 2




CT, Silvia Dedu (Bucharest University of Economic Studies, RO)A general information measure approach to estimating longevity risk and modelling lifetime data

CT, Alexandru Sin (National Institute for Economic Research “Costin C. Kiritescu” Romanian Academy, RO)An integrated risk measure and information measure approach to assessing the inequality of financial aids for agriculture and rural development


CT, Erika Roldán Roa (The Ohio State University, US)Evolution of the homology and related geometric properties of the Eden Growth Model

CT, Daniel Antonio Flores Agreda (Monash University, AU)Bootstrapping clustered data via a Weighted Laplace Approximation


9:30 – 11:30 Thematic Session

TS, Applied Probability and Statistics Organizer: Liliana Blanco Castañeda (Universidad Nacional de Colombia, CO)

Marco Avella, (Columbia University New York, US)Privacy-preserving parametric inference: a case for robust statistics

Martha Tatiana Pamela Jiménez Valderrama, (Universidad de La Salle Bogotá, CO) Comparison of results between fuzzy sets and classical classification, such as clustering methods of family agricultural production units in Colombia


Natalia Hernández Vargas, (Universidad del Norte Barranquilla, CO)Group Elastic Net: towards sparse geometric data analysis for categorical data

Francisco Javier Rodríguez-Cortés (Universidad Nacional de Colombia sede Medellín, CO)The main recent contributions to the analysis of spatio-temporal point patterns

Monday, December 2




CT, Yeliz Mert Kantar (Eskisehir Technical University, TR)Spatial analysis of the traffic accident statistics in the provinces of Turkey: ESDA and Spatial Panel Approaches

CT, İlhan Usta (Eskisehir Technical University, Science Faculty; Department of Statistics, Eskişehir, TR)Estimation of parameters of the geometric process with weibull distribution via ABC Algorithm


12:00 – 13:00Semi Plenary Talk

SPT, Fernando Quintana, (Pontificia Universidad Católica de Chile, CL)On some classes of random partition models and their applications

Random partition models have become quite popular, with many applications in fields such as classification, clustering and change-point detection. This talk will briefly discuss some of these models and their applications, including the case where the notion of repulsion among clusters is explicitly incorporated as a mechanism to encourage parsimonious representations. Throughout, connections with nonparametric Bayesian models will be highlighted.

13:00 – 14:30 Lunch time


TS, New trends in Rice Formula. (In memory of Mario Wschebor) Organizers: Jorge R. León Ramos (Universidad Central de Venezuela, VE) & Jean-Marc Azaïs (Université de Toulouse, FR)

Mario Wschebor (1939-2011) was born in Uruguay and there, also in Venezuela, he left an important trace in the development of mathematics. His main area of research was the theory of stochastic processes. Being the theory of Gaussian processes and fields a subject in which he obtained important achievements. Particular mention deserves his study of the level sets of such processes. It is important to highlight the various generalizations he made of the Kac-Rice formula. Such a formula allows to obtain, among other things, the moments of the measure of the level set. In the last period of his career, he addressed problems related to the determination of the moments of the number of roots and the moments of the measure of nodal

Monday, December 2


sets for random polynomial systems. Also, he tackled the study of delicate problems in algorithmic complexity. His work opened important research ways that we will present in this session.

Céline Delmas (INRA, FR)Mean number and correlation function of critical points of isotropic Gaussian fields

Federico Dalmao (UDELAR, UY)Rice formula: finiteness of moments and jumps

Yohann De Castro (Université d’Orsay, FR)Testing Gaussian Process with applications to super-resolution


Hall 1, CCU-UADY


CS, Stochastic models on Biology Organizer: Arno Siri-Jegousse (IIMAS-UNAM, MX)

For decades now, random models applied to biology have taken an important place in the worldwide probability community. Random models can apply to a wide range of life science thematics such as evolution, population genetics, ecology... On the other side, a large amount of probabilistic tools have been used to study such phenomena and biological applications have motivated the advanced study of such thematics as measure-valued processes, branching and coagulation processes, random partitions, stochastic differential equations... In Mexico, a contributed session with this thematic was organized by A. Wakolbinger during the SPA 2011. Since this time the presence of biological applications to probability has been permanent. The aim of this proposal is to keep on informing the Latin American community about the area, by inviting young and confirmed researchers from various regions of the globe, providing some diverse visions of the thematic.

Matthias Steinrücken (University of Chicago, US)Inferring demographic histories using coalescent hidden Markov Models

Verónica Miró Piña (IIMAS, UNAM, MX)The symmetric coalescent

Sarah Penington (University of Bath, UK)Genealogies in pushed waves

Adrián González-Casanova, (IMUNAM, MX)The Wright-Fisher model with efficiency


Monday, December 2


Hall 2, CCU-UADY14:30 – 16:30Contributed Talks

CT, Miriam Ramírez García (IMUNAM, MX)On feller boundary conditions for spectrally positive Lévy Processes in {R_+}

CT, Aditya Maheshwari (Indian Institute of Management Indore, IN)Time-changed Poisson processes of order k

CT, Tommaso Lando (University of Bergamo, IT and VSB-TU Ostrava, Czech Republic, CZ) A generalized faily of transformed stochastic orders

CT, Alejandro Santoyo Cano (IMUNAM, MX)Local time for stable processes


Hall 3, CCU-UADY


CS, Random dynamical systems wiht jumps II Organizer: Michael Hoegele (Universidad de los Andes, CL)

Random events occur in nature throughout our everyday experiences. Taking stochastic effects into account is of central importance for the development of mathematical models of complex phenomena under uncertainty arising in applications. Macroscopic models in the form of differential equations for these systems contain randomness in many ways, such as stochastic forcing, uncertain parameters, random sources or inputs, and random initial and boundary conditions. The theory of random dynamical systems and stochastic differential equations provides fundamental ideas and tools for the modeling, analysis, and prediction of complex phenomena. The aim of this contributed session is to showcast new developments on the theory of random dynamical systems whose driven noise has a jump structure. The three speakers on this session will deal with different aspects of stochastic differential equations driven by Lévy processes. This is the second part of the session Random dynamical systems with jumps.

Paulo Ruffino (UNICAMP, BR)Bifurcations of stochastic $n$-point motions

André Oliveira Gomes (UNICAMP, BR)Large Deviations for Lévy processes

Joaquín Fontbona (Universidad de Chile, CL)Synchronization of stochastic mean field networks of Hodgkin-Huxley neurons with noisy channels

Michael Hoegele, (Universidad de los Andes, CO) The first passage problem for stable linear delay equations perturbed by power law Lévy noise

Monday, December 2





TS, Inequalities in Probability and Statistics and Boundary Crossing Organizer: Víctor de la Peña (Columbia University, US)

Philip Ernst (Rice University, US)Optimal Real-time detection of a drifting Brownian coordinate Cindy Rush (Columbia University, US) Some new concentration results for Pseudo-Lipschitz Loss Functions

Demissie Alemayehu (Columbia University, US)On the bias and variance of the Odds Ratio, Relative Risk and False Discovery Proportion

Víctor de la Peña, (Columbia University, US)Decoupling and the price of independence




CS, Workshop on Stochastic Control and Games II Organizer: Héctor Jasso Fuentes (CINVESTAV, MX)

The aim of this session is to give a general view on the state-of-the-art in the areas of stochastic dynamic optimization and dynamic games. In particular, it enables researchers, academics, and students to shape a general picture of novel results, models, and applications with a large social impact. This series of talks is part of the annual event organized by the Mexican School of Stochastic Control.

Adolfo Minjarez Sosa (Universidad de Sonora, MX)Zero-sum Markov games in systems of interacting objects

T2, Héctor Jasso Fuentes (CINVESTAV-IPN, MX)Markov-Feller processes with controlled discontinuities

T3, Oscar Vega Amaya, (Universidad de Sonora, MX),Recent results on Markov decision processes with average cost criterion

Oscar Vega Amaya, (Universidad de Sonora, MX)Recent results on Markov decision processes with average cost criterion

Monday, December 2




14:30 – 16:30Contributed Talks

CT, Elizabeth González-Estrada (Colegio de Postgraduados, MX)Shapiro-Wilk test for skew normal distributions

CT, Lizbeth Naranjo Albarrán (UNAM, MX)Parameter estimation of the generalized Pareto distribution for normal baseline distribution

CT, Addy Bolívar Cimé (Universidad Juárez Autónoma de Tabasco, MX) Binary discrimination methods for High Dimensional Data with a geometric representation

CT Jorge Ignacio González Cázares (University of Warwick, UK)Geometrically convergent simulation of the extrema of Lévy Processes



14:30 – 15:30 Contributed talks

CT, Héctor Olivero (Universidad de Valparaíso, CL)An hypothesis test to detect divergent stochastic simulations

CT, Jimmy A. Corzo S. (Universidad Nacional de Colombia, CO)Exact and asymptotic runs test for symmetry

CT, L. Leticia Ramírez (CIMAT, MX)Modeling and Inference for the allostatic load and the occurrence of adverse life experiences in suicide individuals



17:00 – 21:00Opening Ceremony

Welcoming cocktail“Vaquería Yucateca”Performance by Ballet Folclórico – Universidad Autónoma de Yucatán

Monday, December 2



8:30 – 10:30 Course

Course, Sudipto Banerjee (University of California Los Angeles, US)Hierarchical Bayesian Modeling and Analysis for Spatial BIG Data

Hall 1, CCU-UADY


CS, Stochastic analysis and biological models Organizers: María Emilia Caballero & Adrián González Casanova (IMUNAM, MX)

We present several applications of stochastic analysis to biological models.

Maite Wilke Berenguer (TU Bochum, DE)Applications of a Chatterjee,-type generator representation

Fernando Cordero (Uni Bielefeld, DE)General selection models: Bernstein duality and (minimal) ancestral structures

Jason Schweinsberg (UCSD, San Diego, US)Branching Brownian motion and populations undergoing selection

Emmanuel Schertzer (Paris 6, FR)Nested Coalescents and transport-coagulation PDE


Hall 2, CCU-UADY


CT, Gabriel Berzunza (Uppsala University, SE)K-cut on Galton-Watson trees

CT, Haruyoshi Tanaka (Wakayama Medical University, JP)Perturbed finite-state Markov systems with holes and Perron complements of Ruelle operators

CT, Héctor Olivero Quinteros (Universidad de Valparaíso, CL)An Euler Scheme for a system of Piecewise Deterministic Markov Processes under mean field interaction

Tuesday, December 3


CT, Hélene Guérin (Université du Québec à Montréal, CA)Longtime behavior of Zigzag processes in interaction


Hall 3, CCU-UADY


CS, Propagation of Chaos Organizer: María Clara Fittipaldi (FC-UNAM, MX)

The propagation of chaos is a central concept of kinetic theory originated with the work of Mark Kac in 1976 about a colliding mean field kinetic gas model. The initial motivation for the subject was to investigate the connection between a detailed and a reduced description of particles’ evolution. In particular, in statistical mechanics, the notion refers to the following situation: starting with a chaotic configuration of a system consisting of a large number N of particles, the evolution due to the interaction between particles will, in general, destroy the independence property. However, if one focuses on the marginal distribution of some given k particles it will be approximately a product-distribution when the number of particles grows to infinity. So in this sense the chaos propagates at any time horizon as the size of the system tends to infinity. This contributed session will be focused on the current advances in the study of propagation of chaos and mean field interacting particle models applied to different areas, such as statistical mechanics, population genetics and queueing theory. This session will also provide a forum for discussion about the connections between the various methods used in these fields.

Joaquin Fontbona (Universidad de Chile, CL)Quantitative propagation of chaos for generalized Kac particle systems

Roberto A. Cortez Milán (Universidad Andrés Bello, CL)Quantitative uniform propagation of chaos for Maxwell molecules

Matthieu Jonckheere (Universidad de Buenos Aires - CONICET, AR)Convergence properties of many parallel servers under power-of-D load balancing

Pablo Groisman (Universidad de Buenos Aires, AR)Fleming-Viot selects the minimal quasi-stationary distribution: The Galton-Watson case


Tuesday, December 3


Hall 1, Gamma Mérida El Castellano


TS, Stochastic control and applications Organizer: Héctor Jasso (CINVESTAV, MX)

Tomás Prieto-Rumeau (Universidad Nacional de Educación a Distancia, ES)Numerical approximations for continuous-time average Markov decision processes

Richard Stockbridge (University of Wisconsin, US)On the Modelling of Uncertain Impulse Control for Continuous Markov Processes

George Yin (Wayne State University, US)Switching Diffusions with Mean-Field Interactions

Erick Treviño (MATE-UNAM, MX)Long term investment




CS, Free Probability Organizer: Víctor Pérez – Abreu (Instituto Politécnico Nacional, MX)

Free probability was developed in the 1980s by Dan-Virgil Voiculescu, whose initial objective was to solve certain problems relating to the structure of Von Neumann´s algebras. It has been a subject of increasing attention in the last two decades. A particular subject of interest is the parallelism between some concepts in classical and free probability. The purpose of this Contributed Session is to provide recent advances in free probability related to subjects in classical probability. Each presentation will have a five-minutes introductory part for an audience of researchers and students working in classical probability and stochastic processes. The subject of the presentations include parallelism between Fourier and Cauchy transforms with application to variance functions, classical and free extreme value laws, classical and free Lévy bases, a classical and free deconvolutions of probability measures. The order of the potential speakers would be as the authors are presented in this proposal below. The speakers are from four different countries, Canada, Denmark, Mexico and USA, and include young and senior researchers.

Tuesday, December 3


Takahiro Hasebe (Hokkaido University, JP)Unimodal distributions in free probability

Jacek Wesolowski (Warsaw Polytechnic, PL)Conditional expectations through Boolean cumulants and subordination - towards a better understanding of the Lukacs property in free probability

J. C. Wang (University of Saskatchen, CA)Extreme values in free probability

Steen Thorbjørnsen (Aarhus University, DK)On Lévy bases in free probability




TS Random discrete structures and statistical mechanics Organizer: Christina Goldschmidt (Oxford University, UK)

Louigi Addario-Berry (McGill University, CA)Hipster random walks

Omer Angel (University of British Columbia, CA)Random graphs on the circle

Will Perkins (University of Illinois at Chicago, US)Algorithms at Low Temperatures

Kavita Ramanan (Brown University, US)Local Dynamics of Interacting Particle Systems on Large Sparse Graphs


Auditorium, CCU – UADY

11:00 – 12:00Semi plenary talk

SPT, Chris Holmes (University of Oxford, UK)Bayesian nonparametric learning through randomized loss functions

We discuss Bayesian nonparametric learning whereby Bayesian nonparametric models are used to train Bayesian parametric models by way of suitably randomized objective (loss) functions. The resulting posterior models exhibit provably better properties than their conventional Bayesian counterparts when the sampling distribution (likelihood function)

Tuesday, December 3


is misspecified. For additive log-likelihoods inference is achieved through posterior sampling obtained by independent optimizations of randomly re-weighted loss-functions, as opposed to Markov chain Monte Carlo sampling. This avoids issues with MCMC such as burn-in and chain dependence, and is highly scalable on modern computer architectures allowing for samples to be drawn in parallel for the price of a single model optimization. We demonstrate the approach on a number of examples including nonparametric learning for Bayesian logistic regression, variational Bayes (VB), mixture models, and random forests. The work has its foundations in the weighted-likelihood bootstrap of Newton and Raftery (1994).

12:00Plenary Talk

Gerard Ben Arous (Courant Institute, USA)Kac-Rice in very high dimensions: from physics to high dimensional statistics and machine learning…,

13:00 – 14:30 Lunch time

14:30 – 15:30Course

Course, Sudipto Banerjee (University of California Los Angeles, US)Hierarchical Bayesian Modeling and Analysis for Spatial BIG Data

Hall 1, CCU-UADY


TS, Recent advances in Bayesian Nonparametrics Organizer: Igor Pruenster (Bocconi University, IT)

Bayesian Nonparametrics is currently one of the most active research streams in Statistics. During the last decade it has experienced an amazing growth in terms of both research output and size of the community. As for the latter, the field now comprises also leading experts in Frequentist Statistics and Probability Theory on the theoretical side and Biostatisticians, Computer Scientists and Engineers on the applied side. This makes Bayesian Nonparametrics a lively and exciting research field. In this session some of the recent developments concerning theory, modeling and computation are showcased.

Long Nguyen (University of Michigan, US)Posterior contraction for mixtures of product distributions

Pierpaolo De Blasi (University of Torino, IT)epsilon-Approximation to the Pitman-Yor process

Andrés Felipe Barrientos (Duke University, US)

Tuesday, December 3


Bayesian inferences on uncertain ranks and orderings

Marta Catalano (Bocconi University, IT)Measures of Dependence in Bayesian Nonparametrics


Hall 2, CCU-UADY


CS, Finance and Actuarial Science Organizer: Adriana Ocejo (University of North Carolina at Charlotte, US),

The general topic is the applications of probability and statistics tools to Finance and Actuarial Science. The session common topics are portfolio and risk management at the interplay between finance and insurance. The session will start with recent advances in investment problems, and then move to topics involving parameter uncertainty and simulation techniques.

Marcos Escobar-Anel (Western University, CA) Recent advances in constrained portfolio optimization with applications to banking and insurance

Hussein Nasralah (Worcester Polytechnic Institute, USA)Portfolio optimization for small time horizons Anne MacKay, (Universite du Quebec a Montreal, CA)Simulating Heston using explicit weak solutions

Jean-Francois Begin (Simon Fraser University, CA)Economic scenario generator and parameter uncertainty: A Bayesian framework


Hall 3, CCU-UADY


CT, Sarai Hernández Torres (University of British Columbia, CA)Scaling limits of uniform spanning trees in three dimensions

CT, Manuel González-Navarrete (Universidad del Bío-Bío, CL)Lack of phase transitions in staggered magnetic systems. A comparison of uniqueness criteria

CT, Andrés Iturriaga (Universidad de Chile, CL)On dependent Dirichlet processes for general Polish spaces

CT, María Fernanda Gil Leyva Villa (IIMAS-UNAM, MX)Random probability measures based on Beta Markov processes

Tuesday, December 3





CS, Workshop on Stochastic Control and Games I Organizer: Alejandra Fonseca Morales (UNAM, MX)

The aim of this session is to give a general view on the state-of-the-art in the areas of stochastic dynamic optimization and dynamic games. In particular, it enables researchers, academics, and students to shape a general picture of novel results, models, and applications with a large social impact. This series of talks is part of the annual event organized by the Mexican School of Stochastic Control.

Alejandra Fonseca Morales (UNAM, MX)Some results on potential games

Héctor Jasso Fuentes (CINVESTAV-IPN, MX)Markov-Feller processes with controlled discontinuities

Carmen Geraldi Higuera Chan (Universidad de Sonora, MX)The Mitra-Wan forestry model analyzed under a mean field optimal control problem

Saúl Mendoza Palacios (Colegio de México, MX)Optimal transport and one-side matching games




CS, Branching structures Organizer: Sandra Palau (IIMAS-UNAM, MX)

The purpose of this session is to highlight recent developments on the topic of branching processes and branching structures. They constitute a fundamental topic in the theory of stochastic processes and are very interesting from the theoretical point of view as well as for its various applications. This area of probability had experienced an increscent interest in recent years in Latin America, and worldwide, since many complex systems such as Gaussian Free Field, Random Planar maps, biological structures, etc; use a lot of techniques which are currently developed in this area. In this session, we propose a diversity of topics related to Galton-Watson in continuous time, Branching Brownian motion, Continuous state branching processes and superprocesses. Two of the speakers are from the region and the other two speakers have been interacting with many researchers in Latin-America.

Tuesday, December 3


Santiago Saglietti (Israel Institute of Technology, IL)A Strong Law of Large Numbers for Super-critical Branching Brownian Motion with Absorption

Maria Clara Fittipaldi (UNAM, MX)Quasi-Stationary Distribution for the discrete-state continuous-time branching process with logistic growth

Simon Harris (University of Auckland, New Zealand)Genealogies of samples from Galton-Watson processes

Janos Englander (University of Colorado, US)Superdiffusions with super-exponential growth




CS, Asymptotic and multiscale analysis of Stochastic biological systems Organizers: Hye-Won Kang; Wasiur R. KhudaBukhsh (University of Maryland at Baltimore County, US); (The Ohio State University, US)

Stochastic modeling is becoming increasingly popular in biological sciences. The ability to account for intrinsic fluctuations and uncertainty in experimental outcomes has been an advantage of stochastic methods. The application of stochastic tools has proven to be tremendously useful in analyzing biological phenomena. The objective of this contributed session is to highlight some of the recent advances in the area of asymptotic and multi-scale analysis of stochastic biological systems´ both at the micro and the macro level. The session will cover a wide range of themes (including applications and techniques) giving a general overview of the field. The topics include new asymptotic results/approximation methods for intracellular processes in the form of laws of large numbers, functional central limit theorems or large deviations principles. Stochastic chemical networks and their associated reflecting diffusion, large/moderate deviations, quasi-steady-state approximations are studied with possible applications to epidemiology and molecular cell biology. An overview of recent statistical inference algorithms for those systems will be given. Arnab Ganguly (Louisiana State University, US)Large deviations for fully-coupled switching diffusion systems

John Fricks (Arizona State University, US)Integrating nanoscale kinetics into motor-cargo transport systems using semi-Markov processes

Daniel Linder (Augusta University, US)Statistical inference in stochastic reaction networks

Tuesday, December 3


Wasiur R. KhudaBukhsh (The Ohio State University, US) Survival dynamical systems for the population-level analysis of epidemics


Auditorium

17:00 – 18:00 Semi Plenary Talk

SPT, Jean Michel Marin (Universidad de Montpellier, FR)Bayesian model choice as a classification problem

Auditorium’s Lobby

18:00 – 19:00 Posters Session

PS1, Jennifer Acuña Larios (Universidad de Costa Rica, CR) Random walk in a random environmentPS2, Rosangela Assumpção (Universidade Tecnológica Federal do Paraná - UTFPR, BR)Influence diagnostics in parameterized linear spatial models PS3, José Betancourt (Institut de Mathématiques de Toulouse, FR) Kriging metamodeling of functional-inputs computer code for coastal flooding hazard assessment

PS4, Karen Alejandra Bojacá Sánchez (University Santo Tomas, CO) Correlation between pairs of point patterns in a Cox process PS5, Michel Cordoba-Perozo (Colombian Institute for Education Assessment, CO) IRT modelling and Levenshtein distance use for detecting wrong linkage between longitudinal survey´s registers; a large scale... PS6, Fermín Eduardo Cardos Vera (Universidad Autónoma de Yucatán, MX) Phase type fitting of scale functions via the EM algorithm PS7, Elsa Cazelles (Centro de Modelamiento Matemático, Uiversidad de Chile, CL) The Wasserstein-Fourier Distance for Time Series PS8, Alejandra Christen (Pontificia Universidad Católica de Valparaíso, CL) Asymptotic behavior of a stochastic epidemic model SI with linear transmission rate

Tuesday, December 3


PS9, Arrigo Coen Coria (FC-UNAM, MX) Multiple changepoint detection in a nonhomogeneous Poisson Process via genetic algorithms

PS10, Isabel Cristina García Arboleda(Pontificia Universidad Javeriana, CO) Change point detection in mean of short memory process PS11, Ana Gómez (University Escuela Colombiana de Ingeniería Julio Garavito) Approach to Metaheuristic algorithms, applying Machine Learning techniques for the analysis of time series PS12, José Benito Hernández C (CIMAT-Monterrey, MX) Implementation of neural networks in time series to generate a portfolio of investment in cryptocurrencies PS13, Adrián Hinojosa (Federal University of Minas Gerais, BR) Exit time for a reaction diffusion model: Case of a one well potential PS14, Lorena León Velasco (CIRAD and Université de Montpellier, FR) Partitioned conditional generalized linear models for analyzing rice diversity PS15, Alejandro Lopera (Statistician and marketing analyst, CO) Analysis of scattered functional data for electromyographic signals PS16, Humberto Martínez-Bautista (CIMAT,Aguascalientes, MX) The extreme values approach is useful to demonstrate trends of tree-ring width series of Douglas-fir (Pseudotsuga menziesii) regionally PS17, Blanca Xóchitl Muñoz Vargas (Benémerita Universidad Autónoma de Puebla, MX) A study of the dropout in the degrees of the FCFM-BUAP through the Survival Analysis PS18, Silvia María Ojeda (FaMAF-Universidad Nacional de Córdoba, AR) On similarity and edge detection in images processing PS19, Luciana Pagliosa Carvalho (Western Paraná State University, BR) Multivariate and circular analyzes to classify two crop years of soybean yield regarding agrometeorological informations PS20, Luciana Pagliosa Carvalho (Western Paraná State University, BR) Management zones using a priori information PS21, Rafael A. Pérez Abreu C (CIMAT, Aguascalientes, MX) Prediction of shrimp size distribution reared inside submersible sea cages PS22, Tania Roa (Universidad de Valparaíso, CL) Parameter estimation for random sampled regression model with long memory noise

Tuesday, December 3


Wednesday, December 4

PS23, Luisa Rodríguez (Saint Tomas University, CO) Multinomial logistic regression to areal data PS24, Heivar Yesid Rodríguez Pinzón (Universidad de Santo Tomás, CO) Cointegrated panel data model for per capita gross domestic product and electric power consumption of 11 South American countries PS27, Miguel Ángel Uribe Opazo (Universidade Estadual do Oeste do Paraná, BR)Reparameterized t-student spatial modelling and diagnostics applied to soybean productivity data PS28, Margoth Adriana Valdivieso Miranda yVíctor Miguel Burbano Pantoja (Universidad Pedagógica y Tecnológica de Colombia, CO) Probability measures, strategy for teaching PS29, Semra Günay Akta (Anadolu University, Eskişehir, TR) Spatial analysis of satisfaction of Asian Airport

PS30 Jonas Arista (University College Dublin, IE)Loop-erased random walks and random matrices

Auditorium – CCU-UADY

8:30 – 9:30Semi Plenary Talk

Poisson processes of objects, Pablo Ferrari (Universidad de Buenos Aires, AR)


CT, Leonardo Videla (Universidad de Valparaíso, CL)Out-of-equilibrium random walks

CT, Gerónimo Uribe, (IMUNAM, MX)On random walks with preferential relocation and (weighted) random recursive trees


Hall 1 – CCU-UADY


CT, Luis Gorostiza (CINVESTAV. MX)Percolation and Random walks on percolation cluster in hierarchical lattices

Tuesday, December 3



Hall 2 – CCU-UADY


CT, Maria Eulalia Vares (Universidade Federal do Rio de Janeiro, BR)Fast-reaction limit for Glauber-Kawasaki dynamics with two components

CT, Mauricio Duarte (Universidad Andrés Bello, CL)Gravitation vs Brownian motion


Hall 3 – CCU-UADY


CT, André Oliveira Gomes (IMECC-UNICAMP, BR)Homogenization of a mean field PDEs: a probabilistic approach

CT, Víctor Hugo Vázquez Guevara (Benemérita Universidad Autónoma de Puebla, MX)On the almost sure central limit theorem for the Elephant Random Walk


Hall 1 – Gamma Mérida El Castellano Hotel


CT, Felix Leopoldo Rios(The royal Institute of Technology, Stockholm, SWE)Bayesian structure learning in decomposable graphical models using sequential Monte Carlo methods

CT, José Benito Hernández C (CIMAT-Monterrey, MX)Study of cryptocurrencies using time series models




CT, Emmanuel Afuecheta (Nnamdi Azikiwe University, PMB 5025, Awka, NG)Flexible models for stock returns based on Student’s t distribution


CT, Carmen Adriana Gheorghe (National Institute for Economic Research “Costin C. Kiritescu”, Romanian Academy, RO)Asymptotic properties of the Nonparametric estimators of the Leimkuhler Curve and Kakwani Index. Applications to income inequality modelling




CT, Biviana Marcela Suárez Sierra (Universidad Nacional de Colombia, CO)Bivariate Weibull accumulated average function for counting the thresholds of two air pollutants in Bogotá

CT, Abhik Ghosh (Indian Statistical Institute, Kolkata, IN)Robust semiparametric inference for polytomous logistic regression with complex survey design




CT, Kazeem Adeleke (Obafemi Awolowo University, Ile-Ife, NI)Estimation of Cure Rate Model from Survival Data of HIV/AIDS Patients Under Antiretroviral Therapy; Bayesian Approach

CT, Claire Delplancke (Universidad de Chile, CL)A scalable stochastic algorithm for passive seismic tomography



11:00 – 12:00 Plenary Talk

PT, Thomas Mountford (EPFL, Switzerland)Invariance principles for Markov Cookie random walks

We take as starting point the recent work of Kosygina and Peterson (who are coworkers in the current project) and consider Markov cookie random walks in one dimension where the cookies at each site are generated by



a Markov process. In the mean zero, recurrent case we show that the walk converges to a Brownian motion perturbed at extrema. The method is a coarse grained application of Ray Knight techniques.


SPT, Michele Guindani (University of California, Irvine, US)Two-group mixture models for multiple testing

Free afternoon


Thursday, December 5 Auditorium – CCU-UADY

8:30 – 10:30Course

Course, Vincent Tassion (ETH Zurich, CH)Sharpness of the phase transition in percolation


Hall 1 – CCU-UADY


CS, Statistical model selectionOrganizer: Florencia Leonardi (Universidade de São Paulo, BR)

Model selection methods are important tools for statistical modeling. While classical statistical analysis focus on methods for analyzing data given a model with fixed dimension, the model selection methods are used to select the most suitable dimensionality, given the data. Different model selection techniques have been proposed in the literature, such as the Akaike information criterion (AIC) (Akaike, 1974) and the Bayesian information criterion (BIC) (Schwarz et al., 1978), that are methods based on the likelihood function. From an information-theoretic perspective, we can mention the Minimum description length (MDL) principle (Hansen


& Yu, 2001). The goal of this section is to present some model selection techniques and recent results for different types of models, like random graphs and networks, stochastic processes, between other.

Andressa Cerqueira (UNICAMP, BR)Learning communities in weighted networks (Joint work with Elizaveta Levina)

Nancy Garcia (UNICAMP, BR)Hidden Markov random field models applied to color homogeneity evaluation in dyed textiles images (joint work with Victor Freguglia and Juliano L. Bicas)

Roberto Imbuzeiro Oliveira (IMPA, BR)Mean-field models for deep neural networks

Florencia Leonardi (USP, BR)Strong structure recovery for partially observed discrete Markov random fields on graphs (joint work with Lara Frondana and Rodrigo R.S. Carvalho)


Hall 2 – CCU-UADY


CT, Joaquín Fontbona (Universidad de Chile, CL)Bayesian learning with Wasserstein barycenters

CT, Zaida Quiroz (Pontificia Universidad Católica del Perú, PE)Bayesian Block Nearest Neighbor Gaussian process for large point-referenced data

CT, José A. Villaseñor (Colegio de Postgraduados, MX)A test of fit for Gaussian stochastic processes


Hall 3 – CCU-UADY


CT, Felipe Tobar (Universidad de Chile, CL)Band-Limited Gaussian Processes: The Sinc Kernel

CT, Lizbeth Naranjo Albarrán (UNAM, MX)A multi-state model addressing ordinal response misclassification for non-decreasing processes

Thursday, December 5


CT, Mario Santana-Cibrian (CONACYT - IMUNAM, MX)Bayesian inference on the dynamics of acute respiratory diseases in a population structured by age

CT, Alan Riva-Palacio (University of Kent, UK)Bayesian non-parametric survival analysis regression




TS, Limit Theorems for Random Processes Organizer: Glauco Valle (Universidade Federal do Rio de Janeiro, BR)

Luiz Renato Fonte (Universidade de São Paulo, BR)Random walks in birth-and-death environments

Leandro P. R. Pimentel (Universidade Federal do Rio de Janeiro, BR)Brownian Aspects of The KPZ Fixed Point

Enrique Guerra (Pontificia Universidad Católica de Chile, CL)A proof of Sznitman’s conjecture about ballistic RWRE

Rodrigo Ribeiro (Pontificia Universidad Católica de Chile, CL)The tree builder Random Walk




CS, Gausian free field and related topics Organizer: Santiago Juan Saglietti (NYU - Shanghai, CN)

The d-dimensional Gaussian free field (GFF) is a d-dimensional-time analog of Brownian motion. Just as Brownian motion is the limit of the simple random walk (when time and space are appropriately scaled), the GFF is the limit of many incrementally varying random functions on d-dimensional grids. Like Brownian motion, it is a simple random object of widespread application and that has attracted great attention recently for its connections with other important topics within the probability community. It plays an important role in statistical physics and the theory of random surfaces, particularly in the case d = 2. It is also a starting point for many constructions in quantum field theory. Recently, strong links between Gaussian Free Fields on finite graphs and local time fields of simple random walks on such graphs have been established and exploited in the study of cover times of finite graphs. Furthermore, the GFF has also intimate connections with Schramm-Lowener evolutions (SLE), which appear in the study of the level lines of the GFF. Finally, the GFF has also deep links with the



study of the vacant set of random interlacements. The aim of this session is to bring together experts on both discrete and continuum versions of the GFF to communicate on the recent developments made in the area and in connection with the other related topics mentioned above.

Oren Louidor (Technion Israel Institute of Technology, IL)Weak convergence for the scaled cover time of the rooted binary tree

Lisa Hartung (University Mainz, DE)High points of a random model of the Riemann-Zeta function and Gaussian multiplicative chaos

Avelio Sepúlveda (Université Lyon 1, FR)On level sets of the two-dimensional Gaussian free field




CS, Random Matrices Organizer: Octavio Arizmendi (CIMAT, MX)

Random Matrix Theory is nowadays a prominent and very active area in probability, with important applications to other areas in science and engineering such as information theory, machine learning, wireless communications, statistical mechanics, number theory and functional analysis. This session plans to bring speakers from different countries in Latin America, United States and Europe, working in diverse aspects of Random Matrix Theory. This include limiting behavior in the edge of the spectrum, local limit laws, non-hermitian models, Free probability and its applications such as neural networks.

José Alexander Ramírez González (Universidad de Costa Rica, CR)Asymptotic behavior of diffusion processes associated to beta-ensembles

Nicholas Cook (Stanford University, US)Convergence to Brown´s spectral measure for polynomials in Ginibre matrices

Camille Male (CNRS, Université de Bordeaux, FR)Traffic independence and Freeness over the diagonal

Mario Alberto Díaz-Torres (CIMAT, MX)Analysis of artificial neural networks: old and new random matrix theory perspectives





11:00 – 12:00Course

Course, Vincent Tassion (ETH Zurich)Sharpness of the phase transition in percolation

Hall 1 – CCU-UADY


CS, Statistical modeling with biomedical applications (Part 1) Organizer: Carolina Euan (King Abdullah University of Science and Technology, SA)

The statistical inference has a crucial impact on biomedical applications. From a medical perspective, statistics appear since designing the clinical trials to analyzing the final output data. Biostatistics is a branch of statistics that specialize in responding to questions and problems in medicine, public health or biology. In this session, we present some examples of the interaction between statistics and medical studies, with a particular emphasis on the challenges, limitations, and open problems.

Tim Ramsay (Ottawa Hospital Research Institute, CA)P-values have no place in non-inferiority clinical trials

F. Javier Rubio (King’s College London, UK)Net survival models for cancer epidemiology: The Good, The Bad, and The Ugly

Hall 2 – CCU-UADY


CT, Elsa Cazelles (CMM - Center for Mathematical Modeling, CL)Statistical properties of regularized barycenters in the Wasserstein space

CT, Michel Cordoba-Perozo (ICFES - Colombian Institute for Education Assessment, CO)Item Response Theory: Methodological proposal for data augmentation in small samples for item parameters estimation



Hall 3 – CCU-UADY


CT, Zitlalli Salas Gutiérrez (CIMAT, MX)Goodness of fit test by projection selection for high dimensional data

CT, Jairo Díaz-Rodríguez (Universidad del Norte, CO)High-dimensional hypothesis testing under sparse and dense alternatives for generalized linear models



CT, Martin Wiegand (University of Manchester, GB)New approaches for Galactic Ellipticity Estimation

CT, Izhar Asael Alonzo Matamoros (Universidad Nacional Autónoma de Honduras, HN)Bayesian inference with heavy tail distribution for the estimation of the touristic expenditure in Honduras



CT, María Inés Armendáriz (Universidad de Buenos Aires, IMAS-Conicet, AR)Gaussian random permutations and the boson point process



CT, Alejandro Hernández Wences (IIMAS-UNAM, MX)The Bolthausen-Sznitman coalescent: combinatorics and genetics

CT, Lizbeth Peñaloza Velasco (IIMAS-UNAM, MX)The shape of a seed bank tree




13:00 – 14:30 Lunch time14:30 – 16:30Thematic Session

TS, Verblunsky coefficients and polynomials in action in probability and statistics Organizer: Fabrice Gamboa (Institute du Mathematique de Toulouse, FR)

This session aims to discuss and give an overview on some results for models arising from the stochastic world obtained by considering orthogonal polynomials. Alain Rouault (Laboratoire de Mathématiques de Versailles, FR)Large deviations for random measures and sum rules

Reda Chhaibi (Institut de Mathématiques de Toulouse, FR)On the circle, Kahane’s Gaussian multiplicative chaos and circular random matrices match

Laure Dumaz (Laboratoire du CEREMADE Dauphine, FR)Localization of the continuous Anderson Hamiltonian in 1-d and its transition

Jérôme Stenger (EDF and Institut de Mathématiques de Toulouse, FR)Optimal Uncertainty Quantification of a risk measurement from a thermal-hydraulic code using Canonical Moments


Hall 1 – CCU-UADY


CS, Statistical modeling with biomedical applications (Part 2) Organizer: Carolina Euan (King Abdullah University of Science and Technology, SA)

Hernando Ombao (King Abdullah University of Science and Technology, SA)Statistical Methods for Analyzing Brain Signals in a Rat Stroke Experiment

Carolina Euan (King Abdullah University of Science and Technology Saudi Arabia, SA)Coherence-based Clustering for Visualization of Brain Connectivity




CT, Christian Fonseca-Mora (Universidad de Costa Rica, CR)Semimartingales and their stochastic calculus on spaces of distributions

CT, Harold A. Moreno Franco (National Research University Higher School of Economics, RUS)Controlled jump-diffusion processes


Hall 2 – CCU-UADY


CT, Abdolnasser Sadeghkhani (ITAM, MX)Application of Bayesian Statistics in Hockey

CT, Juan Kalemkerian (Universidad de la República, UY)An Independence Test Based on Recurrence Rates

CT, Oscar Peralta Gutiérrez (The University of Adelaide, AU)Rate of strong convergence of stochastic fluid processes to Markov-modulated Brownian motion

CT, Manuel González-Navarrete (Universidad del Bío-Bío, CL)Discrete-time dependent processes with memory lapses


Hall 3 – CCU-UADY


CT, Jairo Díaz-Rodríguez (Universidad del Norte, CO)Estimating the emissivity of a galaxy cluster: A high-dimensional Poisson inverse problemCT, Avram Sorin (National Institute for Economic Research “Costin C. Kiritescu”, Romanian Academy and University of Craiova, RO)Multidimensional models for assessing the impact of urban growth poles on economic convergence under the perspective of regional policies CT, Tommaso Lando (University of Bergamo, IT and VSB-TU Ostrava, Czech Republic, CZ) A generalized faily of transformed stochastic ordersCT, Osvaldo Angtuncio Hernández (IMUNAM, MX)Dini derivatives for exchangeable increment processes and applications CT, Arab Idir, (CMUC, University of Coimbra, PT)Recent results on stochastic convex transform order






TS, Recent advances in Bayesian modelling based on exponential and conjugate families of distributions Organizer: Eduardo Gutiérrez-Peña (IIMAS-UNAM, MX)

Exponential families are an important class of probability models that occur, in one form or another, as part of more complex models widely used in applied statistics (such as generalized linear models, hierarchical models and dynamic models). On the other hand, conjugate families play an important role in the Bayesian approach to parametric inference. One of the main features of such families is that they are closed under sampling, but a conjugate family often provides prior distributions which are tractable in various other respects. In this session we shall discuss some recent advances in statistical modelling and analysis based on exponential and conjugate families of distributions.

Luis Enrique Nieto-Barajas (ITAM, MX)General dependence structures for exponential family models

Manuel Mendoza (ITAM, MX)Bayesian robustness revisited

Eduardo Gutiérrez-Peña (IIMAS-UNAM, MX)Families of multivariate distributions derived from conjugate exponential family models

Alberto Contreras-Cristán (IIMAS-UNAM, MX)Dirichlet process mixtures of von Mises distributions




CS, Statistical Inference for Complex Data Organizer: Ricardo Fraiman (Universidad de la República, UY)

In recent years the amount, the diversity and specially the complexity of the data has considerably increased. This new flow of data, coming from diverse sources like social networks, biochemistry, health care systems, politics, have given rise to new well known and closely related terms like ?Big-Data?, ?manifold learning? and ?complex data?. They can be an image, a video, a sub-graph on a network, a functional data among other, which make it difficult to link, match, cleanse and transform data across systems with traditional methods. Data on unstructured metric spaces. Typically statistical



approach that perform reasonably well for small and well structured will fail since in most of them there is no an Euclidean structure on the space. Some examples are the space of graphs, feature spaces, manifolds, or Banach spaces. To handle them new mathematical and computational methods are needed. Detection, estimation and testing problems will be addressed in this session. In particular, Manifold learning, which is the result of the confluence of at least three classical theories, (a) directional (or circular) data where the aims are similar to those of the classical statistics but the data are supposed to be drawn on the sphere or, more generally, on a lower-dimensional manifold; (b) the study of non-linear methods of dimension reduction, aiming at recovering a lower-dimensional structure from random points taken around it, and (c) some techniques of stochastic geometry and set estimation whose purpose is to estimate some relevant quantities of a set (or the set itself ) from the information.Alejandro Cholaquidis (Universidad de la República, UY)Stochastic detection of some topological and geometric features

Leonardo Moreno (Universidad de la República, UY)Depths on manifolds

Manuel Febrero Bande (Universidad de Santiago de Compostela, ES)TBA

Ricardo Fraiman (Universidad de la República, UY)Detection and identification for complex data




CT, Wayne Nelson (Private consultant, US)Statistical comparison of sets of recurrence data

CT, Luis Ramón Munive Hernández (Universidad Autónoma de Chapingo, MX)Spatio-temporal Point Process Analysis of Forest Fires in Mexico State

CT, Ramsés Mena Chávez (IIMAS-UNAM, MX) Modal posterior clustering motivated by Hopfield’s network






CS, New developments on time-depended functional data Organizer: Israel Martínez Hernández (King Abdullah University of Science and Technology, SA)

Functional Data Analysis (FDA) has been studied intensively over the past ten years, due to its great applicability to a large number of different fields such as finance, climatology, medicine, oceanography, etc. FDA is an alternative approach to deal with high dimensional data, temporal data, and spatial structure. FDA combines mathematics, probability, statistics, and computer science, which allow us to build multi-disciplinary research. In this session, novel models and methods for functional data with possible time-dependence are introduced. First, two model based forecasting methods for functional time series are discussed. Then, two new nonparametric methodologies are presented, one for detecting clusters of functional data, and another for detecting extreme curves. Diego Rivera García (Coppel S.A. de C.V, MX) A proposal for robust clustering of time series

Israel Martínez Hernández (King Abdullah University of Science and Technology, SA)Nonparametric trend estimation in functional time series

Kimihiro Noguchi (Western Washington University, US)Forecasting Intraday Volatility Curves Using Singular Spectrum Analysis

Greg Rice (University of Waterloo, CA)Inference for the autocovariance of a functional time series, and Goodness-of-Fit tests for fGARCH models



18:00SLAPEM Assembly

Conference DinnerEx-Hacienda Ya-aska

Kei Noba (Kyoto University, JP); JL Pérez (CIMAT, MX), X. Yu (The Hong Kong Polytechnic University, CN)On the bail-out dividend problem for spectrally negative Markov additive models

Hiroshi Tsukada (Kyoto University, JP)Pathwise uniqueness of SDEs driven by Cauchy processes with drift



Hall 3, CCU-UADY


CS, Bayesian Uncertainty Quantification in systems of differential equations models Organizer: J Andrés Christen (CIMAT, MX)

In the study of models arising from systems of ordinary or partial differential equations usually there are unknown parameters that need to be inferred from data. This is commonly known as Inverse Problems and in the past few years it has been accepted as a statistical problem. Moreover, providing a probabilistic, Bayesian, solution to Inverse Problems has been recognized as most suited to quantify the uncertainty in our inferences [1, 2]. This broadly refers to Bayesian UQ and has seen a strong growth in interest from large numbers of groups in academia and in industry, with applications in many fields of science and technology.



SPT, Lea Popovic (Concordia University, CA)A spatial measure-valued model for interactions in heterogeneous (intra-cellular) systems

Hall 1, CCU-UADY


CS, Yet more applications of the theory of one-dimensional Levy ProcessOrganizer: Kouji Yano (Kyoto University, JP)

The theory of Levy processes has been deeply studied and extensively applied. The aim of this session is to present a rich variety of new applications of the theory of Levy processes, especially in the one-dimensional case.

Kosuke Yamato, (Kyoto University, JP)Fluctuation scaling limits for positive recurrent jumping-in diffusions

Toru Sera (Kyoto University, JP)Distributional limit theorems for intermittent maps


Friday, December 6


There are, none the less, a series of issues in this Bayesian inference problem that make it specially difficult and particularly challenging, both mathematically and computationally speaking.

A particular difficulty separating this inference problem from others is that the regressor model, eg. the PDE model, commonly cannot be computed directly but a numerical approximation is needed. Such numerical method involves an error and also involves heavy computing times, as it is the case with FEM methods for solving PDEs. This has two clear consequences: first, these numerical errors may be propagated into the posterior distribution and, second, the evaluation of the posterior can be very CPU costly. An other important aspect of Bayesian UQ is that the definition of the prior distribution is commonly non-standard and complex since in many cases the quantities of interest are functions that themselves need to be discretized and given a prior. Moreover, the resulting theoretical posterior distribution is only represented numerically in our computers through a discretization and we require consistency results assuring that the posterior is well defined and the corresponding discretization leads to a numeric posterior with sufficient error control. Last, but not least, we also require very efficient Monte Carlo methods to simulate from the posterior, taking into consideration the discretization error, the multidimensionality and the highly CPU costly evaluation of the posterior.

In this session we will present an introduction to this multidisciplinary subject with a first overview talk and also overview talks on research topics in Bayesian UQ in numerical analysis of PDEs, discretization error control, modeling, prior distributions and Monte Carlo methods, dimensionality reduction, several examples and one case study from Bayesian inversion methods for seismic data in the pacific shore of Mexico.

[1] Kaipio, Jari, Somersalo, E. (2005), ``Statistical and Computational Inverse Problems’’, Springer.[2] Fox, C and Haario, H and Christen, J A (2013) ``Inverse problems’’, In: Bayesian Theory and Applications, Chapter 13, Ed: Damien, P et al, Oxford University Press.

Eduardo Gutiérrez-Peña (IIMAS-UNAM, MX)Overview of Bayesian Uncertainty Quantification

Antonio Capella (IMUNAM, MX)FEM in Bayesian UQ


Marcos Capistrán (CIMAT, MX)Posterior error control in Bayesian UQ

J. Cricelio Montesinos (CIMAT, MX)Bayesian UQ in geophysics applications in Mexico

Friday, December 6



9:30 – 12:00 Thematic Session

CS, Stochastic processes in random environment Organizer: Juan Carlos Pardo Millán (CIMAT, MX) The field of stochastic processes in random environment has been the object of intensive mathematical research over the last thirty years. It covers a variety of models, mainly from condensed matter physics, physical chemistry, biology and geology. The extra randomness can cause very unexpected effects in the large scale behavior of many of these models; on occasions these run contrary to the prevailing intuition in this session we want to present recent developments on two different types of processes in random media; random walks and branching processes. All members of this session are young talented researchers, three speakers are from the region (Brazil, Chile and Mexico) and one speaker from Europe who has been collaborating with latin-americans.Manuel Cabezas (Pontificia Universidad Católica de Chile, CL)The ant in the labyrinth

Marcelo Hilario (Federal University of Minas Gerais, BR)Random walks on conservative interacting particle systems


Sandra Palau (IIMAS-UNAM, MX)Branching processes in Markov additive environment


12:00 – 13:00Aranda – Ordaz Award Session IStatistics

13:00 – 14:30 Lunch time

14:30 – 15:30Aranda – Ordaz Award Session IIProbability

15:30 – 16:30Plenary Talk

Estimating the interraction functions and the graph of interactions in multivariate Hawkes processes using Bayesian nonparametric methods, Judith Rousseau (University of Oxford, UK)

Friday, December 6


Organizing Committee

Addy Bolívar Cimé, UJAT

Daniel Hernández Hernández, CIMAT (Chair)

Gabriel Rodríguez Yam, UACh

Héctor Jasso Fuentes, CINVESTAV

Henry G. Pantí-Trejo, UADY

Joaquín Ortega Sánchez, CIMAT

Jorge A. Argáez Sosa, UADY

José Alfredo López Mimbela, CIMAT

José Luis Batún Cutz, UADY

Juan Carlos Pardo Millán, CIMAT

Leticia Ramírez-Ramírez, CIMAT

Manuel Mendoza Ramírez, ITAM

Miguel Nakamura Savoy, CIMAT

Pedro Uribe Flores, CIMAT

Ramsés Mena Chávez, IIMAS, UNAM

Roberto Quezada Batalla, UAM-Iztapalapa

Sergio I. López Ortega, Facultad de Ciencias, UNAM

Víctor Rivero Mercado, CIMAT

Conference SecretaryRosy Dávalos, CIMAT

r o s a @ c i m a t . m x

Sponsors

TABLE OF CONTENTS - CIMAT

Documents

Transcript of TABLE OF CONTENTS - CIMAT