We will talk about the graph embedding paradigm for tackling machine learning tasks in this type of structures and their implementation in Julia.
Embedding approaches have been proved to be very successful in several Natural Language Processing tasks. Language itself can be thought of as a graph structure where each word can be represented as a vertex and the edges between them are their transition probability.
With this reasoning algorithms such as Word2Vec and current state of the art convolutional neural network techniques have been ported and proposed to the graph world in order to tackle questions like node and graph classification, link prediction and spreading processes forecasting.
We will discuss this techniques and their basic implementation in Julia.