In this talk, I will give an overview of NASA’s work applying machine learning to aviation safety. The machine learning problems include anomaly detection, precursor identification, and text mining. The types of data include Flight Operations Quality Assurance (FOQA) data, radar track (trajectory) data, surface data, and aviation safety reports. We will discuss several insights obtained on each of these types of data that were previously unknown.

Download the slides for this talk.Download ( PDF, 5050.97 MB)