The Pareto Frontier of Resilient Jet Tagging
This addresses potential uncertainty and bias in high-energy collider physics analyses, but appears incremental as it focuses on trade-offs rather than introducing a new method.
The paper tackles the problem of model-dependent bias in jet tagging classifiers by exploring trade-offs between performance metrics and resilience, demonstrating that networks with high performance can have low resilience.
Classifying hadronic jets using their constituents' kinematic information is a critical task in modern high-energy collider physics. Often, classifiers are designed by targeting the best performance using metrics such as accuracy, AUC, or rejection rates. However, the use of a single metric can lead to the use of architectures that are more model-dependent than competitive alternatives, leading to potential uncertainty and bias in analysis. We explore such trade-offs and demonstrate the consequences of using networks with high performance metrics but low resilience.