Machine Learning on Graph-Structured Data
Part 01 – 8:30 – 10:00
Presentation video: https://youtu.be/KC5avDYKNHc
Part 02 – 10:30 – 12:00
Presentation video: https://youtu.be/0bIDxWRkNdg
Several real-world complex systems have graph-structured data, including social networks, biological networks, and knowledge graphs. A continuous increase in the quantity and quality of these graphs demands learning models to unlock the potential of this data and execute tasks, including node classification, graph classification, and link prediction. This tutorial presents machine learning on graphs, focusing on how representation learning — from traditional approaches (e.g., matrix factorization and random walks) to deep neural architectures — fosters carrying out those tasks. We also introduce representation learning over dynamic and knowledge graphs. Lastly, we discuss open problems concerning scalability, robustness, and multidimensional graphs.
- Claudio D. T. Barros
- Daniel N. R. da Silva
- Fabio Porto (LNCC)
Claudio D. T. Barros has been a PhD candidate in Computational Modeling at the National Laboratory for Scientific Computing (LNCC) since October 2017. In 2015, he received a B.Sc. Degree in Nanotechnology with Emphasis on Physics, followed by a M.Sc. Degree in Physics in 2017, all obtained at the Federal University of Rio de Janeiro (UFRJ). He developed a research in Parameter Estimation in Dynamical Systems, focused mainly on Biological Systems and advanced his expertise in Metaheuristics, such as Genetic Algorithms and Differential Evolution. His current research focus on Machine Learning applied to Network Science, in particular Embedding Methods for Dynamic and Multidimensional Networks.
Daniel N. R. da Silva is currently a PhD student in Computational Science at the Brazilian National Laboratory for Scientific Computing (LNCC). He holds an M.Sc. degree in Computational Science from LNCC (2017) and a B.Sc. degree in Computer Science from UNIFESO (2014). His research interests lie at the intersection among Knowledge Engineering, Machine Learning, and Network Science. Daniel works on two related research projects: his PhD thesis development and a model management R&D project. In his thesis, Daniel is interested in the analysis and development of machine learning methods for modelling temporal knowledge graph dynamics, for instance, the inference of a new fact based on the graph information. On the other hand, in the R&D project, Daniel is involved in building a knowledge-based system for the management of machine learning models.
Fabio Porto is a Senior Researcher at the National Laboratory of Scientific Computing, in Brazil. He is the founder of the DEXL Laboratory, developing R&D activities in the context of scientific data analysis and management. He holds a PhD in Informatics from PUC-Rio, with sandwich at INRIA, in 2001, and a postdoc at Ecole Polytechnique Fédérale de Lausanne (EPFL). He has more than 80 research papers published in International Conferences and Scientific Journals, including VLDB, SIGMOD and ICDE. He was the General Chair of VLDB 2018 and SBBD 2015. Since 2018 he has been a member of the SBBD steering committee, and a member of SBC and ACM.