MEM_HEP

Collection of work for my Master's Thesis.

Includes

1. code for (the namegiving) Matrix Element Method for $ZHH$ and $ZZH$ to $l\bar{l}(Z) + b\bar{b} + b\bar{b}$ including the phase-space parameterization for MC integration
   • in VEGAS
   • with a neural importance sampling strategy using rational quadratic coupling transforms by Durkan et al. [1] implemented in the normflows package by Stimper et al. [2]. The integration procedure is based on i-flow [3].
2. the JetConv Marlin processor for jet clustering with GNNs and Spectral Clustering. see GraphJet

Setup

Using mamba and miniforge is recommended. Create a virtual environments

mamba create -n py311 python=3.11 && mamba activate py311

Requires

mamba install pytorch numpy matplotlib seaborn pandas click
pip install vegas uproot

Alternatively, consider conda create --name py311 --file requirements.txt (however, this will install many modules more than what is strictly required).

To build the integrand, check analysis/cffi and

For neural importane sampling, install the forked normflows package

git clone https://github.com/nVentis/normalizing-flows.git
pip install -e normalizing-flows

Using and converting LCIO files requires pyLCIO to be available in the conda environment. See here for build instructions.

How-To

For the integration, check the CLI command python cli mem_integrate and mem_integrate in analysis/mem.py.

Citations

[1] "Neural Spline Flows", by Durkan et al. (2019). Proceedings of the 33rd International Conference on Neural Information Processing Systems (NeurIPS 2019) arXiv:1906.04032

[2] "normflows: A PyTorch Package for Normalizing Flows", by Stimper et al. (2023). Journal of Open Source Software, 8(86), 5361, DOI arXiv:2302.12014.

[3] "i-flow: High-dimensional Integration and Sampling with Normalizing Flows", by Christina Gao, Joshua Isaacson, Claudius Krause (2020). Mach. Learn.: Sci. Technol. DOI arXiv:2001.05486.