Machine Learning algorithms for the classification of blockchain addresses.


This is the second in the series of workshops, dedicated to applying Artificial Intelligence to the blockchain.

The workshop is dedicated to the investigation of how machine learning algorithms can be applied for the classification purposes of blockchain addresses and smart contracts.

We will try to solve the problem of identifying the type of address in the blockchain, based on the transaction history and the interactions with other addresses. Solving this problem has a value in the security-related analytics, optimization of the user experience of applications for the blockchain, finding new trends and entities in the blockchain.

Learn on basic entities of the blockchain and their properties.
Build the features and the dataset, querying the real blockchain data.
Make the machine learning model, and train it on the data set.
Generate the results out of the model, classify the addresses from the data.
Discuss applications of this solution.

Workshop implies initial programming and data analytical skills from the participants, as we will write code, run and analyze the data.
No special knowledge of blockchain technology is required, but nice to have.

We will use only the cloud services, no software except browser will be needed to accomplish the workshop.

It is advisable to take the notebook, as then you can repeat the workshop yourself and make more investigations during the workshop and later.