Towards the Development of Entropy-Based Anomaly Detection in an Astrophysics Simulation

TL;DR

This paper explores the application of entropy-based machine learning anomaly detection techniques to astrophysics simulations, specifically core-collapse supernovae, aiming to improve simulation accuracy and identify anomalies.

Contribution

It introduces a novel approach combining entropy measures with ML for anomaly detection in astrophysics simulations, highlighting strategies, early successes, and future challenges.

Findings

Initial anomaly detection results show promise in identifying simulation irregularities.

Entropy-based methods can enhance the detection of anomalies in complex astrophysical data.

Challenges include computational complexity and the need for tailored models.

Abstract

The use of AI and ML for scientific applications is currently a very exciting and dynamic field. Much of this excitement for HPC has focused on ML applications whose analysis and classification generate very large numbers of flops. Others seek to replace scientific simulations with data-driven surrogate models. But another important use case lies in the combination application of ML to improve simulation accuracy. To that end, we present an anomaly problem which arises from a core-collapse supernovae simulation. We discuss strategies and early successes in applying anomaly detection techniques from machine learning to this scientific simulation, as well as current challenges and future possibilities.