[ABE-L] Convite: Seminário DEST/UFMG em 01/07/2022.

[Vinicius Mayrink]

Caros,

Na próxima sexta-feira (*01 de julho*, às *13:30h*) o ciclo de Seminários
do *Departamento de Estatística da UFMG* terá a apresentação de *Esther
Salazar*.

Esther é pesquisadora do FDA (Food and Drug Administration) nos EUA, e
obteve o grau de Mestre e Doutora em Estatística pela UFRJ. Ela tem
experiência de pesquisa como pós-doc no Statistical and Applied
Mathematical Sciences Institute (SAMSI, EUA) e no Electrical and Computer
Engineering Department (Duke University, EUA). Além disso, já foi docente
do Electrical and Computer Engineering Department (Duke University). Suas
áreas de pesquisa são: Bayesian analysis, survey data analysis, population
modeling, biostatistics, public health, spatio-temporal modeling,
nonparametric methods, computational statistics e machine learning
techniques for big data.

O seminário será transmitido ao vivo pelo canal do Youtube "*Seminários
DEST - UFMG <https://www.youtube.com/channel/UCoZC2_pME9ca_-Hx4djd60w>*".

At.te,
Vinícius Mayrink


*********** Título e Resumo ***********
Esther Salazar (FDA - Food and Drug Administration, EUA)


*Flexible models for heterogeneous multiview data: applications to
behavioral and fMRI data.*
We present a probabilistic framework for learning with heterogeneous
multiview data where some views are given as ordinal, binary, or
real-valued feature matrices, and some views as similarity matrices. Our
framework has the following distinguishing aspects: (i) a unified latent
factor model for integrating information from diverse feature (ordinal,
binary, real) and similarity-based views, and predicting the missing data
in each view, leveraging view correlations; (ii) seamless adaptation to
binary/multiclass classification where data consists of multiple feature
and/or similarity-based views; and (iii) an efficient, variational
inference algorithm which is especially flexible in modeling the views with
ordinal-valued data (by learning the cutpoints for the ordinal data), and
extends naturally to streaming data settings. Our framework subsumes
methods such as multiview learning and multiple kernel learning as special
cases. We demonstrate the effectiveness of our framework on several
real-world and benchmark datasets.

-- 
*Vinícius D. Mayrink*
*Professor Associado - Departamento de Estatística*

*ICEx, Universidade Federal de Minas Gerais*
-------------- Pr�xima Parte ----------
Um anexo em HTML foi limpo...
URL: <http://lists.ime.usp.br/pipermail/abe/attachments/20220624/61accfb0/attachment.htm>