Conferencista: Celina Miraglia Herrera de Figueiredo (UFRJ)

Coordenadora: Dayse Haime Pastore
Local: Anfiteatro do Bloco 3Q
Data e horário: Segunda-feira 16/09 – 17h00

Resumo: "Resolver ou Verificar?" é uma pergunta que vale um milhão de dólares. No ano 2000, o Instituto Clay para Matemática distinguiu sete problemas considerados centrais para o progresso da matemática, chamando-os de Os Problemas do Milênio. A solução de cada problema corresponde a um prêmio de um milhão de dólares. Um dos sete problemas selecionados é um problema de Teoria da Computação: existe pergunta cuja resposta pode ser verificada rapidamente mas cuja resposta requer muito tempo para ser encontrada? Esse Problema do Milênio, conhecido como P versus NP, é o problema central na área de Complexidade Computacional, onde tentamos classificar a dificuldade dos problemas de acordo com a eficiência das possíveis soluções através de algoritmos computacionais.

Conferencista: Márcio A. Murad (LNCC)

Coordenador: Fabricio Simeoni de Sousa
Local: Anfiteatro do Bloco 3Q
Data e horário: terça-feira 17/09 – 10h30

Resumo: We present a new class of multi-scale computational models developed by our group at LNCC to describe multi-phase flow  (water, oil and gas) in carbonate petroleum reservoirs characterized the presence of complex geological features.
Among the geological structures, we particularly highlight  the presence  of fractures, capable of  sustaining  non-linear elastic stresses, and karst conduits (caves, vugs) which arise from calcite dissolution by water with low pH.  A new class of computational methods is developed to discretize the governing equations resulting from the multi-scale modeling. Numerical simulations of hydrocarbon flow in the giant Brazilian Pre-Sals reservoirs are presented to validate the proposed model.

Conferencista: George Corliss (Marquette University)

Coordenador: Weldon Lodwick
Local: Anfiteatro do Bloco 3Q
Data e horário: quarta-feira 18/09 – 10h30

Resumo: Malaria and other mosquito-borne diseases kill about 1 M people each year, mostly children in Africa.  To reduce (and eventually eliminate) malaria, we need to understand the lifecycle of mosquitoes.  Eggs become larvae, which become pupae, which emerge as adults.  Adult female mosquitoes acquire a malaria-causing parasite by biting an infected person.  She rests to digest her blood meal, lays eggs, and returns to transmit the parasite to a new victim.  We discuss a simple, linear compartment model of that lifecycle and advance to model a model of autodissemination (mosquitoes carry minute quantities of insecticide to their breeding sites) and machine learning and artificial neural network models of estimating the age of a mosquito from near-infrared spectra.  We care how old the mosquito is because older mosquitoes are more likely to be carriers of the parasite.

Conferencista: João Frederico C. A. Meyer (Unicamp)

Coordenadora: Rosana Sueli Da Motta Jafelice
Local: Parque do Sabiá (ver no menú Informações Úteis as orientações e informação para chegar no Parque usando nosso transporte)
Data e horário: quinta-feira 19/09 – 18h15

Resumo: Nesta palestra, e usando conceitos básicos de matemática, serão apresentados alguns modelos que descrevam o impacto ambiental em corpos aquáticos e, em seguida, serão mencionados trabalhos existentes em Matemática Aplicada que incluam, além do impacto, possíveis efeitos em diversas diferentes populações. Na parte final, serão descritos alguns modelos que possibilitam contribuir na previsão de dinâmicas no caso de ruptura de barragens. O objetivo é o de enfatizar a necessidade de se ter a Matemática Aplicada e Computacional como instrumento auxiliar na consideração de riscos, desastres, recuperação - e efeitos a longo prazo.

Conferencista: Cláudio T. Silva (New York University)

Coordenador:  Luis Gustavo Nonato
Local: Anfiteatro do Bloco 3Q
Data e horário: sexta-feira 20/09 – 11h30

Resumo: The large volumes of urban data, along with vastly increased computing power, open up new opportunities to better understand cities. Encouraging success stories show that data can be leveraged to make operations more efficient, inform policies and planning, and improve the quality of life for residents. However, analyzing urban data often requires a staggering amount of work, from identifying relevant data sets, cleaning and integrating them, to performing exploratory analyses and creating predictive models that take into account spatio-temporal processes. Visual analytics systems can greatly help in the analysis of urban data allowing domain experts from academia and city governments better understand cities. In this talk, we
discuss our work in building a visual analytics framework to interactively explore large spatio-temporal data sets and give an overview of our research that combines visualization and data management to tackle these challenges.

Conferencista: Cláudia Linhares Sales (UFC)

Coordenador: Carlos Hoppen
Local: Anfiteatro do Bloco 3Q
Data e horário: terça-feira 17/09 – 19h00

Resumo: With a wide number of applications, the coloring problem of graphs is a NP-hard problem. One way of dealing with it is by the mean of heuristics, aiming to obtain a good coloring while being sure about their worst performance. In this talk, we will present some heuristics to color graphs, with a special attention to b-coloring and greedy coloring. We start by discussing complexity and upper bounds of their worst performance. Then, we recall some structural properties of graphs which directly impact the heuristics, such as girth, maximum degree, m-degree and forbidden subgraphs. We finish this talk by showing two recently defined b-colorings parameters, namely b-homomorphism and partial b-coloring and some of their open problems of general interest.

Conferencista: Leo Dorst (University of Amsterdam)

Coordenador: Carlile Campos Lavor
Local: Anfiteatro do Bloco 5S
Data e horário: terça-feira 17/09 – 19h00

Resumo: Over the past decades, it is becoming clear that Geometric Algebra provides a suitable extension and even replacement of linear algebra in its geometric (or geometrized) applications. Both at the conceptual, computational and implementation levels, it is simply a much better foundation of the structure of geometrical relationships and operators.

GA computes directly with the geometric elements themselves rather than through their coordinates. By incorporating subspaces computationally, and by its unlimited choice of appropriate metrics, it enormously extends the number of primitives - as well as automatically encoding their relationships. It has universal operators to move (or transform) all its elements, independent of their dimension, and parameterized naturally. We are routinely using them to replace matrices in our applications.

I provide an introductory overview of how GA accomplishes this, interactively illustrated with specific examples. I will also indicate some mathematical challenges that arise out of this novel structure, in the hope that the audience can help address them.