Open Problems in Robotic Anomaly Detection
Abstract
Contents include: Background, Anomaly Detection, Abstract, Non-malicious faults present many false alarms, When is invalid data anomalous?, How do we update our assumptions?, Intentional anomalous behavior and emergency stops, Hierarchies of systems with shared functionality, Distribution of computation across hosts, Distributed and Efficient ML, Fixing anomalies on the fly.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 31, 2018
- Accession Number
- AD1084364
Entities
Organizations
- Carnegie Mellon University