<div dir="ltr"><div>Boa tarde a todos(as),</div><div><br></div><div>O tradicional Ciclo de Seminários do Programa de Pós-graduação em Estatística da UFPE inicia suas atividades com transmissão online via <b>Google Meet<span><span><span><span><span><span><span></span></span></span></span></span></span></span></b>. <br></div><div><br></div><div><div><span><span><span><span>A agenda completa de seminários confirmados pode ser acessada pelo site: <br></span></span></span></span></div><div><span><span><span><span><br></span></span></span></span></div><div><span><span><span><span><a href="https://sites.google.com/view/seminarios-ppge-ufpe" target="_blank">https://sites.google.com/view/seminarios-ppge-ufpe</a>. <br></span></span></span></span></div><div><span><span><span><span><br></span></span></span></span></div><div><span><span><span><span>Segue a informação da próxima palestra (<b><span><span><span><span><span><span><span></span></span></span></span></span></span></span></b><span><span><span><span><span><span><span>o link será divulgado na página</span></span></span></span></span></span></span> no dia do seminário): </span></span></span></span><br><span><span><span><span></span></span></span></span></div><span><span><span><span></span></span></span></span></div><div><span><span><span><span><b><br></b></span></span></span></span></div><div><span><span><span><span><b>Dia e hora: </b>17/06 (16h00)<br></span></span></span></span></div><div><span><span><span><span><b><br></b></span></span></span></span></div><div><span><span><span><span><b>Título:</b> Flexible distribution-free conditional predictive bands.<br><br><b>Palestrante:</b> Rafael Izbicki - UFSCar. <br><br><b>Resumo:</b>
Conformal methods create prediction bands that control average coverage
assuming solely i.i.d. data. Besides average coverage, one might also
desire to control conditional coverage, that is, coverage for every new
testing point. However, without strong assumptions, conditional coverage
is unachievable. Given this limitation, the literature has focused on
methods with asymptotic conditional coverage. In order to obtain this
property, these methods require strong conditions on the dependence
between the target variable and the features. We introduce two conformal
methods based on conditional density estimators that do not depend on
this type of assumption to obtain asymptotic conditional coverage:
Dist-split and CD-split. While Dist-split asymptotically obtains optimal
intervals, which are easier to interpret than general regions, CD-split
obtains optimal size regions, which are smaller than intervals.
CD-split also obtains local coverage by creating prediction bands
locally on a partition of the features space. This partition is
data-driven and scales to high-dimensional settings. In a wide variety
of simulated scenarios, our methods have better control of conditional
coverage and have a smaller length than previously proposed methods.<br><br></span></span></span></span><div><span><span><span><span><b>Sobre o palestrante:</b>
Bacharel e Mestre em Estatística pela Universidade de São Paulo, Rafael
é PhD em Estatística pela Carnegie Mellon University (2014). Atualmente
é Professor Adjunto da UFSCar- Universidade Federal de São Carlos. Tem
experiência na área de Probabilidade e Estatística, com ênfase em
Machine Learning (aprendizado de máquina), Bioestatística,
Astroestatística, Fundamentos da Estatística, Inferência Bayesiana,
Inferência Não Paramétrica e Inferência em Dados com Alta
Dimensionalidade. Mais informações: <a href="https://www.google.com/url?q=http%3A%2F%2Fwww.rizbicki.ufscar.br&sa=D&ust=1592418067100000&usg=AOvVaw3p7m-zBK3CJgpJ0RLJEerh" target="_blank">http://www.rizbicki.ufscar.br</a> (Fonte: Currículo Lattes). <br></span></span></span></span></div><div><span><span><span><span><br></span></span></span></span></div><div><span><span><span><span>Favor de avisar a possíveis interessados(as)!</span></span></span></span></div><div><span><span><span><span><br></span></span></span></span></div><div><span><span><span><span>Um abraço</span></span></span></span></div><div><span><span><span><span>Pablo M. Rodriguez<br></span></span></span></span></div></div></div>