Defense Event

Anomaly Detection and Sequential Filtering with Partial Observations

Elizabeth Hou

Friday, July 19, 2019
3:00pm - 5:00pm
EECS 1005

Add to Google Calendar

About the Event

Abstract: With the rise of "big data" where any and all data is collected, comes a series of new challenges involving the computation and analysis of such massive data sets. Nowadays, data is continuously collected leading to questions of at which point should analysis begin and how to incorporate new data into the analysis. Additionally, within the massive amounts of data collected, the features of interest may not be directly observed or only a small subset of the data with certain properties may be of interest. And while collecting data is "cheap", labeling data often is not. In this talk, I will touch upon models for data with partial labels, latent variables, and anomalies in scenarios where such data is continuously being collected. I will also outline some real world applications including cyber security, transportation, and weather systems.

Additional Information

Sponsor(s): Professor Alfred Hero

Open to: Public