Quantum algorithms and machine learning for high-energy physics modelling

High-energy physics (HEP) involves the generation and processing of immense amounts of data; consequently, modern HEP often deploys ML methods to parse abundances of data and extract signal. Yet, even with state-of-the-art ML and deep learning techniques, it remains challenging for HEP researchers to efficiently analyse the masses of experimental data.

Meanwhile, quantum information theory promises a new paradigm for computation. Quantum algorithms promise to achieve advantages in resource costs over classical algorithms, and thus push beyond the limits of classical computation. ML is no exception; quantum ML (QML) seeks to explore the potential for quantum computers to surpass the performance of classical approaches in big data tasks.

Consequently, this raises the question: Can quantum algorithms and QML provide novel tools for future HEP analyses?

This project will contribute to answering this question, by developing new approaches to QML and quantum stochastic modelling, and evaluating their performance on HEP datasets.

The main aim is to explore the potential present and future utility of QML in HEP research. The three core objectives are as follows:

1) Develop new approaches to QML, with particular focus on methods applicable to HEP, e.g., enhancing anomaly detection for discovery of beyond Standard Model physics.

2) Develop applications of memory-efficient quantum models of stochastic processes and adaptive systems for stochastic sampling problems and beyond, e.g., for enhanced rare event sampling.

3) Benchmark these new proposals against current classical techniques and other existing quantum algorithms and QML approaches, particularly in a HEP context, using experimental HEP data sets.