O MDCC está recebendo a visita do Prof. Dennis Shasha, renomado cientista da Ciência da Computação da Universidade de Nova Iorque, no período de 22/02 a 02/03/2016.

Segue um pequeno resumo de algumas de suas habilidades:

"Professor Dennis Shasha works on quite a few different projects. Most have to do with large data and pattern matching or machine learning. Areas of interest include computational biology and biomedicine (data analysis, visualization, experimental design), time series (fast algorithms for fundamental problems such as correlation and burst detection as well as applications like query by humming), and pattern matching in trees and graphs. Beginning in 2013, Ihe has been intrigued by the problems and opportunities in millimeter wireless under the auspices of NYU Wireless. This has included projects having to do with making magnetic resonance image reconstruction faster and core topics like channel modeling and databases for propagation measurements. A general pattern? He likes puzzles. A second general pattern is that he programs a lot in a fast, extremely expressive language called K and its successor languages like q."


O professor proferirá duas palestras, conforme abaixo:

Palestra #1: The Changing Nature of Invention in Computer Science
Local/Horário: Sala de Seminários Bloco 952, Terça-Feira (23/Fevereiro) às 9:00h

What drives inventions in computing? Necessity seems to play only a minor role. Anger at the way things are is much more powerful, because it leads to easier ways to work (the invention of new computer languages). A general dissatisfaction with the practical or theoretical structure of the world can open up whole new approaches to problems (complexity theory and cryptography). Finally, a genuine collaboration between people and machines can lead to an entirely new kind of engineering for devices that will travel to far-off planets or to hostile environments. The talk will discuss the work of several inventors in computing and engineering, their inventions, and how they came up with them and how they plan to come up with more in the future. The ensuing discussion will address the fundamental nature of invention in a world partly populated by intelligent machines.


Palestra #2: Fast Calculations of Simple Primitives in Time Series
Local/Horário: Sala de Seminários Bloco 952, Quarta-Feira (24/Fevereiro) às 9:00h

Financial time series streams are watched closely by millions of traders. What exactly do they look for and how can we help them do it faster? Physicists study the time series emerging from their sensors. The same question holds for them. Musicians produce time series. This talk presents techniques and case studies for three problems:

i. Finding sliding window correlations in financial, physical, and
other applications.
ii. Discovering bursts in large sensor data of gamma rays.
iii. Matching hums to recorded music, even when people don't hum well.

This tutorial draws mostly from the book  High Performance Discovery in Time Series: techniques and case studies, Springer-Verlag 2004 but will discuss some techniques that post-date the publication of that book.

Convidamos a todos os alunos e professores a se fazerem presentes.