Designing Observables for Measurements with Deep Learning
This work addresses the challenge of improving measurement precision in particle and nuclear physics by replacing heuristic-based observable design with a data-driven method, offering incremental advancements in simulation-based analyses.
The paper tackles the problem of designing observables for parameter inference in particle physics by proposing a machine learning approach that trains neural networks with a custom loss function to create targeted observables, resulting in increased sensitivity for distinguishing physics models and reduced unfolding uncertainty in simulations.
Many analyses in particle and nuclear physics use simulations to infer fundamental, effective, or phenomenological parameters of the underlying physics models. When the inference is performed with unfolded cross sections, the observables are designed using physics intuition and heuristics. We propose to design targeted observables with machine learning. Unfolded, differential cross sections in a neural network output contain the most information about parameters of interest and can be well-measured by construction. The networks are trained using a custom loss function that rewards outputs that are sensitive to the parameter(s) of interest while simultaneously penalizing outputs that are different between particle-level and detector-level (to minimize detector distortions). We demonstrate this idea in simulation using two physics models for inclusive measurements in deep inelastic scattering. We find that the new approach is more sensitive than classical observables at distinguishing the two models and also has a reduced unfolding uncertainty due to the reduced detector distortions.