BIP: Boost Invariant Polynomials for Efficient Jet Tagging
This work addresses computational inefficiency and lack of interpretability in jet tagging algorithms for high-energy physics, which is crucial due to large data volumes, but it is incremental as it builds on existing physics-inspired methods.
The authors tackled the problem of inefficient and non-interpretable deep learning architectures for jet tagging in high-energy physics by introducing a novel framework that is invariant to Lorentz boosts, achieving high accuracy and being orders of magnitude faster in training and evaluation.
Deep Learning approaches are becoming the go-to methods for data analysis in High Energy Physics (HEP). Nonetheless, most physics-inspired modern architectures are computationally inefficient and lack interpretability. This is especially the case with jet tagging algorithms, where computational efficiency is crucial considering the large amounts of data produced by modern particle detectors. In this work, we present a novel, versatile and transparent framework for jet representation; invariant to Lorentz group boosts, which achieves high accuracy on jet tagging benchmarks while being orders of magnitudes faster to train and evaluate than other modern approaches for both supervised and unsupervised schemes.