We apply the graph embedding from LINE on crime prediction, and get an MRE of roughly 0.4 on crime prediction. The question is whether this embedding is reliable.
Solution: clustering on embedding vector
Cluster regions based on their embeddings, then explore whether it makes sense.
Problem: Is this graph embedding meaningful?
We apply the graph embedding from LINE on crime prediction, and get an MRE of roughly 0.4 on crime prediction. The question is whether this embedding is reliable.
Solution: clustering on embedding vector
Cluster regions based on their embeddings, then explore whether it makes sense.