A0553
Title: Topological data analysis of dynamic Ethereum token networks
Authors: Yuzhou Chen - Princeton University (United States) [presenting]
Abstract: Forecasting price in the dynamic Ethereum token networks data is indispensable for understanding the blockchain dynamics and measuring the risk connectedness among the cross-cryptocurrency trades. In the last few years, Geometric Deep Learning (GDL), e.g., Graph Convolutional Networks (GCNs), have emerged as a powerful alternative to more conventional time-series predictive models. Despite their proven success, GCNs tend to be limited in their ability to simultaneously infer latent temporal relations among entities. We make the first step on a path of bridging the two emerging directions, namely, time-aware GDL with time-conditioned topological representations of complex dynamic Ethereum token networks. To summarize such time-conditioned topological properties, we develop novel topological representations. We then propose topology-based GDL models which allow us to simultaneously learn co-evolving intra- and inter-dependencies in the dynamic Ethereum token networks data.
