skip to content
Home

Microstructural Kinetics Group

Department of Materials Science & Metallurgy
 

School of Physical Sciences

Subscribe to School of Physical Sciences feed
This is a superlist combining all those seminars on talks.cam taking place in one of the Departments of the School of Physical sciences, plus occasional other talks which would be of significant interest to researchers in the School. If you would like your talk or list included please contact Duncan (drs45)
URL: http://talks.cam.ac.uk/show/index/6793
Updated: 46 min 46 sec ago

Mon 03 Nov 12:00: Title TBC OFBW75 - BIDs4Tech: Quantum Simulation and Quantum Computing

Wed, 13/08/2025 - 11:30
Title TBC

OFBW75 - BIDs4Tech: Quantum Simulation and Quantum Computing

Add to your calendar or Include in your list

Mon 03 Nov 10:55: Title TBC OFBW75 - BIDs4Tech: Quantum Simulation and Quantum Computing

Wed, 13/08/2025 - 11:30
Title TBC

OFBW75 - BIDs4Tech: Quantum Simulation and Quantum Computing

Add to your calendar or Include in your list

Mon 03 Nov 10:20: On Quantum Extended Church-Turing Thesis OFBW75 - BIDs4Tech: Quantum Simulation and Quantum Computing

Wed, 13/08/2025 - 11:30
On Quantum Extended Church-Turing Thesis

OFBW75 - BIDs4Tech: Quantum Simulation and Quantum Computing

Add to your calendar or Include in your list

Mon 03 Nov 10:00: Welcome and Introduction OFBW75 - BIDs4Tech: Quantum Simulation and Quantum Computing

Wed, 13/08/2025 - 11:30
Welcome and Introduction

OFBW75 - BIDs4Tech: Quantum Simulation and Quantum Computing

Add to your calendar or Include in your list

Thu 04 Sep 14:00: Virtual retractions in graphs of groups and applications OGGW02 - Actions on graphs and metric spaces

Wed, 13/08/2025 - 11:30
Virtual retractions in graphs of groups and applications

A subgroup H of a group G is a virtual retract if H is a retract of a finite index subgroup K in G, that is, H RFRS groups (in the sense of Agol). In the talk I will discuss necessary and sufficient criteria for G to have (VRC), where G is the fundamental group of a finite graph of groups. In the case when the vertex groups are virtually abelian these criteria can be checked using elementary tools from Linear Algebra and Euclidean Geometry. I will also discuss applications of (VRC) to algebraic and geometric properties of G. The talk will be based on recent joint work with Jon Merladet

OGGW02 - Actions on graphs and metric spaces

Add to your calendar or Include in your list

Tue 07 Oct 19:15: All models are wrong and yours are useless: making clinical prediction models impactful for patients

Wed, 13/08/2025 - 10:38
All models are wrong and yours are useless: making clinical prediction models impactful for patients

Most published clinical prediction models are never used in clinical practice which leads to a huge gap between academic research and clinical implementation. Here, I propose a checklist to enable academic researchers to be proactive partners in improving clinical practice and to design models in ways that ultimately benefit patients. Over the years I have come to see academic papers not as ends in themselves, but as the beginning of the journey to clinical implementation, and I am frustrated with how little of my own work ever had clinical impact. I argue that you should outline the road to implementation whilst designing your prediction tool by thinking about what medical decisions they are making and how these tools can be used in routine practice. I will illustrate these ideas by discussing AI models from our work to find a minimally invasive alternative to endoscopy and looking at ways of assessing breast cancer survival rates after surgery that can be used in the clinic.

Add to your calendar or Include in your list

Tue 03 Feb 19:15: Connecting the False Discovery Rate to shrunk estimates

Wed, 13/08/2025 - 10:37
Connecting the False Discovery Rate to shrunk estimates

Science is currently facing a ‘replication crisis’ – a concern that many scientific findings reported are difficult or impossible to reproduce. A major cause of this is the availability of technology that permits the exploration and testing of very large numbers of hypotheses, some of which will almost certainly show large or significant effects by chance, even when no real effects are present: this is the ‘multiplicity’ or ‘multiple testing’ problem. The tools available to address this problem include: shrunk estimates, which reduce the estimated effect in relation to each hypothesis from the observed value towards the null value, and the False Discovery Rate (FDR), which relates to the subset of the hypotheses tested for which the discovery of an effect is announced, and states the proportion of these ‘discoveries’ that is expected to be false. This talk will first examine the conceptual basis for each of these tools, then consider how they are connected. Though shrunk estimates and the FDR are both conventionally presented in the frequentist statistical framework, they can both also be presented in empirical-Bayesian terms, with the prior distribution being provided by: the distribution of effect sizes over the full set of hypotheses (in the case of shrunk estimates), and the distribution of significance-test p-values over the subset of hypotheses giving ‘discoveries’ (in the case of the FDR ). Based on this connection, a formal relationship between shrunk estimates and FDR values, for a normally-distributed response variable, will be illustrated. The talk will conclude by considering which of the two tools is the more appropriate in different practical circumstances.

Add to your calendar or Include in your list

Wed 13 Aug 10:00: Welcome and Introduction OFBW74 - Uncertainty in Machine Learning: Challenges and Opportunities

Wed, 13/08/2025 - 09:30
Welcome and Introduction

OFBW74 - Uncertainty in Machine Learning: Challenges and Opportunities

Add to your calendar or Include in your list

Tue 09 Sep 10:30: Break / Group Photo BIDW01 - The Physics and Mathematics of Boundaries, Impurities, and Defects

Tue, 12/08/2025 - 15:30
Break / Group Photo

BIDW01 - The Physics and Mathematics of Boundaries, Impurities, and Defects

Add to your calendar or Include in your list

Mon 01 Sep 15:30: The geometry of Hitchin grafting representations OGGW02 - Actions on graphs and metric spaces

Tue, 12/08/2025 - 15:30
The geometry of Hitchin grafting representations

Misha Kapovich in collaboration with B. Leeb and J. Porti and also Dey developed a geometric theory of Anosov representations into simple Lie groups of higher rank. Among others, they discovered the importance of studying such representations using invariant Finsler metrics.  Best known such representations are Hitchin surface group representations into SL(n,R). For so-called Hitchin grafting representations we discuss a different geometric viewpoint which can be thought of as a geometric interpretation of Fock Goncharov positivity. It can be used to understand the geometry of these representations explicity. This is based on joint work with Pierre-Louis Blayac and Theo Marty. 

OGGW02 - Actions on graphs and metric spaces

Add to your calendar or Include in your list

Wed 05 Nov 15:00: tbc

Tue, 12/08/2025 - 12:01
tbc

Abstract not available

Add to your calendar or Include in your list

Thu 21 Aug 14:45: Self-organized dynamics in living materials DNMW06 - Recent challenges in the mathematical design of new materials

Mon, 11/08/2025 - 21:30
Self-organized dynamics in living materials

Living systems—from cellular assemblies to animal groups—exhibit remarkable self-organization driven by internal activity and physical interactions. In this talk, we explore how active forces shape large-scale behaviors in living materials, focusing on three representative systems:  bacterial elastica, living chiral crystals, and worm collectives. By integrating theory, simulations, and experiments, we show how synthetically designed as well as natural interactions can lead to interesting emergent properties such as programmable pattern formation, collective oscillations, and ultra-fast response dynamics.

DNMW06 - Recent challenges in the mathematical design of new materials

Add to your calendar or Include in your list

Wed 13 Aug 15:40: Learning the Physics: AI for Electric Motor Design and Evaluation OFBW74 - Uncertainty in Machine Learning: Challenges and Opportunities

Mon, 11/08/2025 - 14:30
Learning the Physics: AI for Electric Motor Design and Evaluation

OFBW74 - Uncertainty in Machine Learning: Challenges and Opportunities

Add to your calendar or Include in your list

Tue 19 Aug 10:30: Nonparametric estimation of trawl processes: Theory and Applications  RCL - Representing, calibrating & leveraging prediction uncertainty from statistics to machine learning

Mon, 11/08/2025 - 09:30
Nonparametric estimation of trawl processes: Theory and Applications 

This talk introduces a flexible class of stochastic processes, called trawl processes, which are defined as Lévy bases evaluated over deterministic trawl sets and are widely applicable in many sciences. We will present a novel nonparametric estimator of the trawl function characterising the trawl set and the serial correlation of the process and establish the corresponding asymptotic theory. A simulation study shows the good finite sample performance of the proposed estimator, and, in an empirical illustration, the new methodology is applied to modelling and forecasting high-frequency financial spread data from a limit order book and to estimating the busy-time distribution of a stochastic queue.   This is joint work with Orimar Sauri (Aalborg University).  

RCL - Representing, calibrating & leveraging prediction uncertainty from statistics to machine learning

Add to your calendar or Include in your list

Thu 28 Aug 16:00: Uncertainty quantification in Gaussian Graphical Models RCLW04 - Early Career Pioneers in Uncertainty Quantification and AI for Science

Fri, 08/08/2025 - 15:30
Uncertainty quantification in Gaussian Graphical Models

Co-authors: Jack Jewson and David Rossell. Gaussian graphical models are widely used to infer dependence structures. Bayesian methods are appealing to quantify uncertainty associated with structural learning, that is on the plausibility of conditional independence statements given the data, and on parameter estimates. However, computational demands have limited their application when the number of variables is large, which prompted the use of pseudo-Bayesian approaches. We propose fully Bayesian algorithms that provably scale well to high dimensions when the data-generating precision matrix is sparse, at a similar cost to the best available pseudo-Bayesian methods. Our examples show that the methods extend the applicability of exact Bayesian inference from roughly one hundred to roughly one thousand variables (equivalently, from 5,000 edges to 500,000 edges) if one desires a solution within a few seconds or minutes. All methods are implemented in the R package mombf.

RCLW04 - Early Career Pioneers in Uncertainty Quantification and AI for Science

Add to your calendar or Include in your list

Wed 13 Aug 10:10: Learning Feasible Sets: Machine Learning for Vehicle Design Optimization OFBW74 - Uncertainty in Machine Learning: Challenges and Opportunities

Fri, 08/08/2025 - 13:30
Learning Feasible Sets: Machine Learning for Vehicle Design Optimization

OFBW74 - Uncertainty in Machine Learning: Challenges and Opportunities

Add to your calendar or Include in your list

Wed 27 Aug 11:00: Physics informed Gaussian process priors RCLW04 - Early Career Pioneers in Uncertainty Quantification and AI for Science

Fri, 08/08/2025 - 12:30
Physics informed Gaussian process priors

Given a linear partial differential equation (PDE), I will discuss the design of Gaussian process (GP) priors for the approximation of its solutions as well as its physical parameters. As standard PDE theory is based on weak or distributional formulations, I will describe necessary and sufficient conditions on the kernel of a centered GP so that its realisations solve the PDE almost surely, in the distributional sense. I will then describe necessary and sufficient conditions so that the samples of a centered GP lie in a given Sobolev space, as the latter are particularly well-suited for the study of PDEs. Importantly, both results do not make any continuity assumptions on the GP model, as PDE theory shows that such assumptions are sometimes too restrictive. If time permits, I will describe an application of such GP models for the approximation of the initial data of the 3D wave equation, as is e.g. the goal of photo-acoustic tomography.

RCLW04 - Early Career Pioneers in Uncertainty Quantification and AI for Science

Add to your calendar or Include in your list

Thu 25 Sep 16:00: Scattering in Field Theory OOEW09 - Strings, CFT and Monopoles: Celebrating Peter Goddard's 80th birthday

Fri, 08/08/2025 - 11:30
Scattering in Field Theory

The scattering polynomials leading to the Cachazo-He-Yuan representation of tree level amplitudes are described both on and off shell, and for individual graphs. The proof these non-Lagrangian amplitudes correspond to conventional gauge field theory scattering uses a recurrence relation, and the conformal Yangian provides symmetry. 

OOEW09 - Strings, CFT and Monopoles: Celebrating Peter Goddard's 80th birthday

Add to your calendar or Include in your list

Thu 28 Aug 14:30: Scalable sequential design for Bayesian inverse problems via conditional transport RCLW04 - Early Career Pioneers in Uncertainty Quantification and AI for Science

Fri, 08/08/2025 - 05:30
Scalable sequential design for Bayesian inverse problems via conditional transport

We present a scalable approach to sequential optimal experimental design for Bayesian inverse problems with expensive forward models and high-dimensional parameters. By combining transport maps, a derivative-based upper bound on expected information gain, and dimension reduction via likelihood-informed subspaces, our method enables tractable experimental design in a sequential setting. We demonstrate the effectiveness of the approach with examples from groundwater flow and photoacoustic imaging.This talk is based on joint work with Tiangang Cui, Roland Herzog, and Robert Scheichl. 

RCLW04 - Early Career Pioneers in Uncertainty Quantification and AI for Science

Add to your calendar or Include in your list

Thu 28 Aug 11:00: Proper Scoring Rules - Estimation and Forecast Evaluation RCLW04 - Early Career Pioneers in Uncertainty Quantification and AI for Science

Thu, 07/08/2025 - 17:30
Proper Scoring Rules - Estimation and Forecast Evaluation

Proper scoring rules have been a subject of growing interest in recent years, not only as tools for evaluation of probabilistic forecasts but also as methods for estimating probability distributions. In this talk, we review the mathematical foundations of proper scoring rules including general characterization results and important families of scoring rules and discuss their role in statistics and machine learning for estimation and forecast evaluation. Based on joint work with Prof. Johanna Ziegel, ETH Zurich.

RCLW04 - Early Career Pioneers in Uncertainty Quantification and AI for Science

Add to your calendar or Include in your list

Contact us

Microstructural Kinetics Group Department of Materials Science & Metallurgy University of Cambridge 27 Charles Babbage Road Cambridge CB3 0FS + (0) 1223 334300

Contact: mkg@msm.cam.ac.uk

Site Privacy & Cookie Policies

Quick Links

Map to the Department of Materials Science & Metallurgy

Department of Materials Science & Metallurgy

University of Cambridge talks

Find us

 

Study at Cambridge

About the University

Research at Cambridge