Back

Speaker "Dongming Liang" Details Back

 

Topic

User Behavior Analytics with Scalable Data Store

Abstract

User behavior analytics is generally used to analyze employees as they may have intentions to steal sensitive information of business. One example can be insider threats detection and prevention. We address this problem by optimizing resources and time with advanced machine learning algorithms for various data volumes. In this talk we will discuss how to extend single PostgreSQL system to meet our scalability requirement. We will present the approaches to data loading, complex SQL queries, and certain constraints enforcement. A cluster of the system that we are using will be demonstrated.


 

Profile

Dongming is a lead software engineer at Capital One Data Intelligence team. He's currently interested in real time stream data processing, machine learning, big data and cyber security systems. He is a committer to Apache Apex project.