The SIDur Exec are pleased to announce the programme for the SIDur PGR Symposium on Friday February 7th from 10:00 in MCS 1022 (VisLab) in the Maths & Computer Science Building.
Programme
10:00 – 10:15 Welcome & Coffees
10:15 – 11:15 Challenges of making quantum computing a reality
Nick Chancellor, Newcastle School of Computing, AMBER
Abstract: In this talk I will review several areas in which I am working related to quantum computing and how they play an important role in advancing the field. This talk will not assume previous quantum expertise, and in fact a central theme will be the importance which quantum non-experts play in making quantum computing a reality and the importance of hybrid quantum/classical techniques.
11:15 – 11:30 Coffee Break
11:30 – 11:50 Leveraging Rust’s Type System for Safe and Efficient Metropolis Simulations
David Lanners, Durham Maths, Pure Group
Abstract: Metropolis simulations are a fundamental tool in statistical physics and computational science, enabling the study of complex systems through stochastic sampling. However, ensuring both flexibility and performance in their implementation can be challenging. This talk explores how Rust’s strong type system facilitates the development of generic and efficient Metropolis algorithms. By leveraging traits, generics, and static dispatch, we enforce correctness while minimising runtime overhead and reducing code duplication.
In this talk, we introduce a computational framework that utilises Rust’s robust type system to implement gauge theories generically, minimizing code repetition. This approach facilitates rapid iteration while maintaining high performance through static dispatch.
11:50 – 12:10 The Surrogate Methodology for Computationally Intensive Models
Zhaocheng Li, Durham Maths, Statistics in Uncertainty Quantification
Abstract: We build simulators/models to mimic and predict the behavior of real-world systems. Complicated systems always come up with computationally intensive models that are expensive or even infeasible to approximate “best” parameters locally or globally. We build a surrogate in terms of Bayesian emulators to balance the usage of numerical method and analytical approach to achieve the feasible computation. Depending on the task we want to handle, this methodology can emulate the simulator’s behavior, perform the global parameter searching, and provide the uncertainty quantification for its inference.
12:10 – 13:30 Lunch
13:30 – 13:50 Nonlinear analysis of the Keller–Segel system with logistic growth
Luci Mullen, Durham Maths, Applied Maths
Abstract: The Keller–Segel system with logistic growth exhibits different types of solution structures in distinct parameter regimes. In this talk, we discuss how liner stability theory can fail to accurately predict the existence of pattern forming instabilities, and instead perform a weakly nonlinear analysis to obtain cubic and quintic amplitude equations. We then use solutions of these equations to predict the shape and amplitude of patterns which linear stability theory fails to predict. Finally, we discuss the spatio-temporal oscillations exhibited in particular parameter regimes and consider the role of competing modes as a possible explanation for this behaviour.
13:50 – 14:10 Pulse Based Implementations of Encoder Circuits in Quantum Machine Learning
Kaalkidan Sahele, Durham Computer Science + Matter Systems Research Group, University of Toronto
Abstract: Quantum computing has been seen to improve on the results that classical computation can produce in various ways due to its ability to exploit concepts in quantum mechanics, such as superposition, entanglement, and parallelism. Quantum machine learning is a rising field in quantum computing that makes use of quantum algorithms to improve the machine learning pipeline. The most common form of quantum machine learning circuits that are used are parameterised quantum circuits – a hybrid of classical and quantum computing, where the regular machine learning pipeline is taken and improved upon by replacing the training component with a quantum computer. In order to use a quantum computer, classical data must be encoded into a quantum state. This is a crucially important step, as the quality of our data must be preserved, and is typically achieved with a gate-based encoder circuit. However, working at the gate level provides limited freedom. This project explores the novel pulse based implementation of encoder circuits, with the beginnings of promising results, and potential future trajectories for this new approach in parameterised quantum machine learning circuits.
14:10 – 14:30 Coffee Break
14:30 – 14:50 Seismic Super-resolution Leveraging Machine Learning Techniques
Mukthar Opeyemi Mahmud, Durham Geophysics, Earth Safe
Abstract: Earth imaging is central to our ability to understand our planet and is important for exploration
for critical minerals and geothermal energy resources, detection, and mitigation of natural
hazards such as earthquakes and the study of plate tectonics. As a result, there is a need for
more precise images of the earth’s interior. However, as this imaging process is ill-posed and
lossy, the images obtained are inevitably a blurry version of the truth. This makes it challenging
to robustly interpret results and draw inferences about geophysical systems.
In our attempt to improve the quality of these images, we will explore opportunities for ‘seismic
super-resolution’: generation of higher-resolution images by combining observed data with
prior knowledge about likely structures and the physics of wave propagation. In order to achieve
this, we draw on ideas from homogenization theory, which aims at computing larger-scale
effective properties of a medium containing small-scale heterogeneities. We will also
explore some machine learning techniques for numerical upscaling, particularly physicsinformed neural networks – which enable us to ensure that the underlying laws of physics are embedded within results.
In this presentation, we will highlight some of the challenges and opportunities in this approach
and present some early results from numerical experiments.
14:50 – 15:10 Scalable non-intrusive UQ workflows in HPC environments
Chung Ming Loi, Durham Computer Science, Scientific Computing
15:10 – 15:30 Closing Remarks & Group Photo