Wanted to do something involving ethereum and machine learning

What it does

Compares ethereum tokens by the graph embedding of their transaction network

How we built it

Uses a weighed sum of node2vec embeddings for graph of a particular token's transaction network to compute embedding for that token. Dataset created by using bigquery to query ethereum public dataset. Cosine similarity to compare graph embeddings.

Challenges we ran into

Processing data

Accomplishments that we're proud of

What we learned

What's next for Ethereum Token Analysis

Built With

Share this project: