Abstract

Analyzing connected data sets such as social networks, electrical grids, and biological networks requires running complex algorithms on graph data structures. We will discuss graph algorithms as simple as breadth first search to advanced algorithms such as community detection, unsupervised clustering, and semisupervised learning with belief propagation. High performance implementations of these algorithms cannot be easily written in traditional high level languages, but can be written in Julia due to novel programming language properties. I’ll speak about recent research in applied graph theory and its relationship to the two language problem and discuss how the JuliaGraphs ecosystem tackles these problems. This enables further expansions of data science research in diverse applications.