Speaker
Description
The analysis of collision events at the Large Hadron Collider (LHC) presents significant computational challenges, particularly due to the need for large amounts of Monte Carlo simulation to reduce statistical uncertainties in the simulated datasets. The most computationally intensive task in Monte Carlo detector simulation is the simulation of high-energy particles interacting with the calorimeter. In this work, we propose a novel approach that combines recent advancements in generative models and quantum annealing techniques to provide fast and efficient simulation of high-energy particle-calorimeter interactions. Our approach, the Quantum Variational Encoder (QVAE), utilizes a Variational Autoencoder (VAE) model with a Restricted Boltzmann Machine (RBM) prior implemented on an annealing Quantum Processing Unit (QPU). The quantum annealing QPU can generate a large number of samples from the latent space of a trained VAE model with high efficiency. We show the performance of the QVAE on simulated calorimetric cluster data. The promising evaluation results demonstrate the accuracy and reliability of our Quantum Variational Encoder. Furthermore, our proposed approach has the potential for significant improvement by extending it to use QPU samples during the training process, enhancing the computational efficiency even further.
| Keyword-1 | Generative Models |
|---|---|
| Keyword-2 | Quantum Computing |
| Keyword-3 | HL-LHC, calorimeters |
