Compiler-Research Monthly Meeting [15 May 2025 at 17:00 CET/8:00PDT]
Vassil Vassilev
Our next meeting will be on 15 May 2025 at 17:00 Geneva (CH) Time .
We are happy to host Jamie Gooding.
Title: Accelerating simulation-based inference in RooFit at LHCb with Clad
Abstract: Simulation-based inference (SBI) is a set of statistical inference approaches in which Machine Learning (ML) algorithms are trained to approximate likelihood ratios, e.g., as an alternative to the likelihood fits commonly performed in HEP analyses. A demonstrator has been developed in which SBI is applied to extract parameters of interest from the kinematic and angular distributions of B0 → D*⁻ μ⁺ νμ decays in pseudodata samples generated with RapidSim representative of the datasets used in LHCb analysis. The SBI fit is constructed using the RooFit framework, to which enhanced Python interfaces were recently introduced. Dense Neural Networks (DNNs) were trained to distinguish between the Standard Model and New Physics scenarios for varying parameters of interest. This workflow incorporates automatic differentiation (AD) of the learned likelihoods, using the ROOT SOFIE framework to generate C++ code from the DNNs, from which gradient code is generated by source-code transformation AD with Clad. In this talk, the SBI fit is introduced and compared to the current state-of-the-art. Additionally, this talk discusses the functionality introduced to enable differentiation from likelihood through to ML classifier call and the impact of this functionality upon performance of the SBI fit.
Speaker: Jamie Gooding is a 3rd year Doctoral
Student at Technische Universitat
(TU) Dortmund and SMARTHEP Early Stage Researcher, working on the
LHCb
experiment and currently completing a secondment with the ROOT
team. His research focuses on developing real-time analysis
techniques
to enable studies of charge-parity and lepton-flavour violation in
neutral B meson decays, with an additional interest in
computational
tools for statistical analyses in HEP. Prior to starting his
doctoral
studies at TU Dortmund, he completed a Physics MPhys (Hons) at the
University of Manchester.
Feel free to forward this email to everybody who might be
interested. This talk will be recorded.
Vassil
Vassil Vassilev
Best, Vassil