Bitcoin Flows Graph Analysis - Neo4j

This project aimed at analysing bitcoin graph data and providing sophisticated information on bitcoin transaction data and aggregate on-chain statistics derived from graph data. The outcome statistics in the form of timeseries were later used in predictive machine learning models.

The following statistics were computed based on graph analysis:

  • daily number of distinct active addresses on the Bitcoin blockchain
  • daily historical supply of coins in blockchain, i.e., sum of BTC paid out in coinbase rewards
  • grouped bitcoin wallets (set of addresses belonging to the same entity) using heuristic clustering deanonymizing transactions
  • wallets’ inflows and outflows, i.e, the amount od incoming and outgoing BTC to/from the wallet

This project was a continuation of my two previous projects:

In this project, I was using the data I modeled in the previous projects to perform graph analysis in Neo4j. I was working with Neo4j Graph Data Science Library as well ad Neo4j parallel processing procedures from the APOC Library.

The statistics I produced within this project were used by financial analysts of an advisory company focusing on investments in cryptoassets, in order to create investment recommendations for their clients.

Data Scientist | Machine Learning Engineer | AI Advisor

20 years of experience in data processing from A to Z