Convite da defesa de dissertação do Programa de Pós-Graduação em Ciência da Computação

A Coordenação do Programa de Pós-Graduação em Ciência da Computação tem a satisfação de convidá-lo para a Defesa de Dissertação:


Learning the Latent Structure of Networked Point Processes

Guilherme Resende Borges


A temporal point process is a sequence of timestamps representing the occurrences of certain events. Recently, there has been a lot of work on the use of Networked Point Processes to extract the latent graphs of large scale networks. The problem can be framed as follows: Given a set of timestamps (e.g. messages or posts) for K entities (e.g. users or web pages), can we extract a latent K by K graph that corresponds to an underlying complex network using only the timestamps? This task has gained attention of researchers and practitioners due to its wide applicability and several models have been proposed to extract such relationships in different datasets. However, we have found that such models are usually under-evaluated. That is, past endeavours focus on a small set of unjustified metrics for one or two datasets at most. To provide the community with a rigorous benchmark, we propose an evaluation framework of metrics and datasets for network inference via Point Processes.


Comissão Examinadora:


Prof. Pedro Olmo Stancioli Vaz de Melo - Orientador (DCC - UFMG)

Prof. Renato Martins Assunção - Coorientador (DCC - UFMG)

Prof. Flavio Vinicius Diniz de Figueiredo - Coorientador (DCC - UFMG)

Prof. Daniel Ratton Figueiredo (PESC - UFRJ)

Prof. Heitor Soares Ramos Filho (DCC- UFMG)


5 de Dezembro de 2019



Sala 2077 do ICEX

Última modificação em Sexta, 29 Novembro 2019 21:33