Combined spatial and temporal Graph Embedding workflows for GCN's

 Suggested by: Johannes Scholz 

Keywords: Graph Convolutional Networks, Spatio-Temporal Graph Embeddings, Transport Processes, Deep Learning, Graph Neural Networks, Temporal Data Modeling 

 

Objective: Develop a workflow for generating combined spatial and temporal graph embeddings for Graph Convolutional Networks (GCNs) to enhance spatio-temporal reasoning and evaluation, particularly for detecting transport processes. 

 

Short Description: This thesis explores the integration of spatial and temporal graph embeddings within Graph Convolutional Networks (GCNs) to improve the detection and understanding of transport processes. By developing a structured embedding workflow, the research aims to enable more effective spatio-temporal reasoning in dynamic systems. The proposed approach will be evaluated on relevant datasets to assess its accuracy in identifying transport phenomena. 

 

Start: 

As soon as possible 

 

Prerequisites/qualification: 

• Knowledge in GIS and spatial data processing 

• Interest/Knowledge in Graph Convolutional Networks