Year of Graduation
Modelling Influence of Shock Events on Country Trade Networks Using Graph Embeddings
Financial Technology and Data Analysis
International trade is an important sphere of economics, which might increase wellbeing of trading countries. International trade depends on relations between countries, which occasionally start trade wars. Economics is interested in predicting how trade volumes change in response to some shocks. Economics apply some trade models, which have their drawbacks. On the other hand, trade can be presented as a graph, and machine learning remarkably advanced in analysis of graph structures. In this work, graph embeddings models are applied with aim to get features of edges in trade graphs, which are able to reconstruct data on weights of exports of each node. It is showed, that these features fail to reconstruct weights. We should continue to try other graph embeddings models.