Rare event modeling with self-regularized normalizing flows: what can we learn from a single failure?

TL;DR

This paper introduces CalNF, a self-regularized normalizing flow framework that effectively models rare failure events from limited data, aiding in understanding and debugging safety-critical autonomous system failures.

Contribution

The paper presents CalNF, a novel regularization approach for normalizing flows that improves modeling of rare events with minimal data, demonstrated on failure modeling and real-world case studies.

Findings

CalNF outperforms existing methods on failure data modeling tasks.

CalNF enables analysis of the root causes of complex failures.

The framework achieves state-of-the-art results in inverse problems with limited data.

Abstract

Increased deployment of autonomous systems in fields like transportation and robotics have seen a corresponding increase in safety-critical failures. These failures can be difficult to model and debug due to the relative lack of data: compared to tens of thousands of examples from normal operations, we may have only seconds of data leading up to the failure. This scarcity makes it challenging to train generative models of rare failure events, as existing methods risk either overfitting to noise in the limited failure dataset or underfitting due to an overly strong prior. We address this challenge with CalNF, or calibrated normalizing flows, a self-regularized framework for posterior learning from limited data. CalNF achieves state-of-the-art performance on data-limited failure modeling and inverse problems and enables a first-of-a-kind case study into the root causes of the 2022 Southwest Airlines scheduling crisis.