Paypal, Inc.

Lichao Liu

Zhanghao Hu

Xuyao Hao

Michael Jim Tien Chan

Novel technical ways of analyzing a blockchain system using machine learning are presented, including structures and techniques that can facilitate blockchain address risk assessment via graph analysis. In various embodiments, a system can access a blockchain. The system can build a transaction graph based on analysis of the blockchain. Nodes of the transaction graph can respectively represent blockchain addresses that are recorded in the blockchain. In various cases, edges in the transaction graph can respectively represent blockchain transactions between different ones of the blockchain addresses that are recorded in the blockchain. The system can calculate risk scores respectively corresponding to the blockchain addresses, based on analyzing the transaction graph via at least one machine learning algorithm. These techniques can improve computer operational efficiency by avoiding the execution of unnecessary blockchain transactions.

Blockchain address risk assessment via graph analysis

A method includes receiving a set of training data and selecting a first machine learning platform based on a first optimization function that metrics past machine learning platforms used for training on the set of training data. The method also includes selecting a first algorithm supported by the first machine learning platform based on a second optimization function that metrics past algorithms used for training on the set of training data. Further, the method includes determining one or more hyperparameters supported by the first algorithm based on a third optimization function that metrics past combinations of hyperparameters from the set of hyperparameters used for training on the set of training data. The method also includes training a machine learning model on the set of training data using the first machine learning platform, the first algorithm, and the one or more hyperparameters.

Hierarchy optimization method for machine learning

A method includes training a first machine learning model based on a set of training data and based on the training, determining a first performance metric corresponding to the first machine learning model. The method also includes determining one or more past performance metrics corresponding to one or more machine learning models that were previously trained based on the set of training data. Based on the first performance metric and the one or more past performance metrics, the method includes automatically selecting a second machine learning model to train based on the set of training data.

Growing community of inventors

Sunnyvale, CA, United States of America

Lichao Liu