Obtaining valuable insights from complex network data

Abstract: Many types of data can be represented by means of relationships between entities: physical/logical connections in computer networks, traffic exchange between devices, individuals on social networks, protein-protein interaction, etc. Hidden among these relationships there is a rich source of features that can provide us with many interesting insights. How can organisations leverage all this information and create new and interesting products based on structured data? The answer to this question lies in the field of Graph Theory, that is, the study of graphs.

Graphs are mathematical structures for the representation of pairwise relationships between objects. Low level features extracted from graphs such as centrality, degree distribution, spectrum, etc. are interesting and informative by themselves, but that's only the very surface of what graph theory can offer. As mathematically elegant as it is, it is also a source of ingenious, elegant and powerful algorithms and tools.

How can a field, which originated at the end of the 19th century, be of any use in today's data analytics world? Recent discoveries and studies in the subfield of complex networks open the door to many diverse applications. A complex network is a graph with non-trivial topological features that we can usually find in real-world systems. Getting familiar with concepts such as network communities, complex network models and dynamic networks will grant us access to a new range of tools to get more value from data. This talk, in which we will share experiences and examples, is aimed at changing your perception about this exciting and interesting growing field.

Bio: Pablo Suau started his career in Academia, where he did research and gave lectures on machine learning and other related topics for 10 years. It was during this time that he had the opportunity to be introduced to the field of Complex Networks. Being attracted by the fast-paced and short-term results oriented industry world, he decided to transition to the private sector four years ago, and he has been working as a Data Scientist since then. He has been in charge of several tasks, including building applied machine learning algorithms and designing new features and solutions based on different types of data.