(Joint work with Coralia Cartis) The problem of finding the most extreme value of a function, also known as global optimization, is a challenging task. The difficulty is associated with the exponential increase in the computational time for a linear increase in the dimension. This is known as the ``curse of dimensionality''. In this talk, we demonstrate that such challenges can be overcome for functions with low effective dimensionality --- functions which are constant along certain linear subspaces. Such functions can often be found in applications, for example, in hyper-parameter optimization for neural networks, heuristic algorithms for combinatorial optimization problems and complex engineering simulations.
We propose the use of random subspace embeddings within a(ny) global minimisation algorithm, extending the approach in Wang et al. (2013). We introduce a new framework, called REGO (Random Embeddings for GO), which transforms the high-dimensional optimization problem into a low-dimensional one. In REGO, a new low-dimensional problem is formulated with bound constraints in the reduced space and solved with any GO solver. Using random matrix theory, we provide probabilistic bounds for the success of REGO, which indicate that this is dependent upon the dimension of the embedded subspace and the intrinsic dimension of the function, but independent of the ambient dimension. Numerical results demonstrate that high success rates can be achieved with only one embedding and that rates are for the most part invariant with respect to the ambient dimension of the problem.
 

Roger Penrose is the ultimate scientific all-rounder.  He started out in algebraic geometry but within a few years had laid the foundations of the modern theory of black holes with his celebrated paper on gravitational collapse. His exploration of foundational questions in relativistic quantum field theory and quantum gravity, based on his twistor theory, had a huge impact on differential geometry. His work has influenced both scientists and artists, notably Dutch graphic artist M. C. Escher.

Roger Penrose is one of the great ambassadors for science. In this lecture and in conversation with mathematician and broadcaster Hannah Fry he will talk about work and career.

This lecture is in partnership with the Science Museum in London where it will take place. Please email @email to register.

You can also watch online:

https://www.facebook.com/OxfordMathematics

https://livestream.com/oxuni/Penrose-Fry

The Oxford Mathematics Public Lectures are generously supported by XTX Markets.

Johann Sebastian Bach was the most mathematical of composers. Oxford Mathematician and Cambridge organ scholar James Sparks will explain just how mathematical and City of London Sinfonia will elaborate with a special performance of the Goldberg Variations. 

---

James Sparks - Bach and the Cosmos (30 minutes)

City of London Sinfonia - J S Bach arr. Sitkovetsky, Goldberg Variations (70 minutes)

Alexandra Wood - Director/Violin

--

Please email @email to register

Watch live:
https://www.facebook.com/OxfordMathematics
https://www.livestream.com/oxuni/Bach-Cosmos

The Oxford Mathematics Public Lectures are generously supported by XTX Markets

We will discuss a basic framework to deal with coherent sheaves on schemes over $\mathbb{Z}$, involving infinite-dimensional results on the geometry of numbers. As an application, we will discuss basic results, old and new, on arithmetic ampleness, such as Serre vanishing, Nakai-Moishezon, and Bertini. This is joint work with Jean-Benoît Bost.

Images are a rich source of beautiful mathematical formalism and analysis. Associated mathematical problems arise in functional and non-smooth analysis, the theory and numerical analysis of partial differential equations, harmonic, stochastic and statistical analysis, and optimisation. Starting with a discussion on the intrinsic structure of images and their mathematical representation, in this talk we will learn about variational models for image analysis and their connection to partial differential equations, and go all the way to the challenges of their mathematical analysis as well as the hurdles for solving these - typically non-smooth - models computationally. The talk is furnished with applications of the introduced models to image de-noising, motion estimation and segmentation, as well as their use in biomedical image reconstruction such as it appears in magnetic resonance imaging.

Consider a network of agents connected by communication links, where each agent holds a real value. The gossip problem consists in estimating the average of the values diffused in the network in a distributed manner. Current techniques for gossiping are designed to deal with worst-case scenarios, which is irrelevant in applications to distributed statistical learning and denoising in sensor networks. We design second-order gossip methods tailor-made for the case where the real values are i.i.d. samples from the same distribution. In some regular network structures, we are able to prove optimality of our methods, and simulations suggest that they are efficient in a wide range of random networks. Our approach of gossip stems from a new acceleration framework using the family of orthogonal polynomials with respect to the spectral measure of the network graph (joint work with Raphaël Berthier, and Pierre Gaillard).