Back

Speaker "Rishiraj Pravahan" Details Back

 

Topic

Machine Learning Use case

Abstract

Anomaly Detection In Time Series Data Time series data is ubiquitous and many applications require machine learning analysis of time series data. Statistical analysis of time series data is an old and well studied subject. However, the advent of the Internet of Things, with sensors that collect and stream time dependent data has necessitated unsupervised and supervised machine learning tools that can run on data collected over time. This talk will survey techniques and tools available to run machine learning algorithms on time series data specifically to identify anomalies while walking through an illustrative example of DNS servers that utilize system level data such as network load, cpu and memory to determine various user loads on DNS servers as well as identify DDoS attacks as anomalies.

Profile

Rishiraj Pravahan is a data scientist working for AT&T. Prior to joining AT&T. Rishiraj worked for the ATLAS experiment at CERN where he was part of the team that discovered the Higgs Boson. While at CERN, he worked on constructing, commissioning and calibrating the ATLAS detector as well as on software techniques to analyze the massive dataset from the Large Hadron Collider to search for new physics. He has also been a passionate teacher and advocate for science through public talks and seminars in the US, Europe, India and Latin America. His current work involves, understanding networks, privacy and security of customer data, collection, storage and analysis of sensor data and making advances in the frontiers of statistics and machine learning. In his spare time he loves to read, play pool, cook, travel and learn about other cultures.