Abstract
Monte Carlo methods are widely used in particle physics to integrate and sample probability distributions on phase space. We present an Artificial Neural Network (ANN) algorithm optimized for this task, and apply it to several examples of relevance for particle physics, including situations with non-trivial features such as sharp resonances and soft/collinear enhancements. Excellent performance has been demonstrated, with the trained ANN achieving unweighting efficiencies between 30% – 75%. In contrast to traditional algorithms, the ANN-based approach does not require that the phase space coordinates be aligned with resonant or other features in the cross section.
Highlights
Monte Carlo (MC) techniques are widely used for sampling multi-dimensional probability distribution functions and for numerical integration in multi-dimensional spaces
The particular application studied in this paper occurs in elementary particle physics, where the outcome of a collision of two particles is described by a pdf, the differential cross section, which can typically be calculated within a Quantum Field Theory framework
Our work demonstrates that the Artificial Neural Network (ANN)-based algorithm is not hobbled by these issues, opening the road to a broader application of such algorithms in particle physics contexts
Summary
Monte Carlo (MC) techniques are widely used for sampling multi-dimensional probability distribution functions (pdfs) and for numerical integration in multi-dimensional spaces. The algorithm iteratively adjusts the edges of the intervals to better approximate the desired pdf This construction can be viewed as map between an input space and a target space. The ANN is a smooth, rather than piecewise linear, map between the input and target spaces, determined by the set of parameters w, which are adjusted to give a good approximation to the target pdf. This provides two major advantages over traditional VEGAS. Appendix A contains details on the relative performance of various ANN architectures
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.