# GROUP LUNCH

## 2015/2016 Term 2 (B1.01, 1:00pm)

*Jan. 4th: *Cancelled.

*Jan. 18th: ***Matt Moores** (UoW), *Bayesian modelling of Raman spectroscopy*

Relevant material: Abstract

*Jan. 25th: *** *Lunch takes place in B3.01*.**

*Feb. 1st: *** Ben Graham** (UoW), * Iterated-integral signature and machine learning*

Relevant material: Abstract

*Feb. 8th: *** *In B3.01* Matt Dunlop** (UoW), *Non-parametric estimation of diffusions and Bayesian posterior contraction rates*

Relevant material: Abstract, Paper1, Paper2

*Feb. 15th: *** Jon Cockayne** (UoW), *Probabilistic Meshless Methods for Bayesian Inverse Problems*

Relevant material: Abstract

*Feb 22nd/23rd/24th: ***EQUIP/AIB meeting** on *Data Assimilation and Inverse Problems.* No EQUIP lunch Feb 22nd

*Feb. 29th: *** Neil Chada** (UoW), * Reduced Order Models for Uncertainty Quantification*

Relevant material: Paper

*Mar. 7th: *** Francois-Xavier Briol** (UoW), *Probabilistic Numerics Approaches to Integration*

Relevant material: Abstract

*Mar. 14th: *** Yulong Lu** (UoW), * Gaussian approximations for transitions paths*

Relevant material: Abstract

**2015/2016 Term 1 (D1.07, 1:00pm)**

Oct. 5th: **Aretha Teckentrup** (UoW), *Dimension-independent convergence rates for Gaussian Process regression*

Relevant material: Paper1, Paper2

*Oct. 12th: ***Daniel Roy** (Department of Statistical Sciences, University of Toronto)*, Computational Barriers to Representing Conditional Independence
*Relevant material: Abstract

*Oct. 19th: ***Jere Koskela** (UoW), * Importance Sampling Rare Trajectories in Reverse Time *

Relevant material: Abstract

*Oct. 26th: ***Daniel Sanz-Alonso** (UoW), * Computing Integrals in High Dimensional Spaces*

Relevant material: Paper1, Paper2, Paper3

*Nov. 2nd: ***Joris Bierkens** (UoW), * MCMC by means piecewise of deterministic sample paths*

Relevant material: Paper

*Nov. 9th: *** Felipe Medina Aguayo** (UoW), *Stability of Noisy Metropolis-Hastings*

Relevant material: Paper

*Nov. 16th: ***Cyril Chimisov** (UoW), * Adaptive Gibbs sampler*

Relevant material: Paper1, Paper2, Paper3

*Nov. 23rd:* ** Matteo Icardi** (UoW),* Quadrature based moment methods for PDF equations and UQ*

Relevant material: Slides, Paper1, Paper2, Paper3, Paper4, Paper5, Paper6, Paper7, Book

*Nov. 30th: *** Shiwei Lan** (UoW), *Dimension-Independent MCMC algorithms for PDE-contrained BIPs*

Relevant material: Abstract

*Dec. 7th: *** Håkon Hoel** (Department of Mathematics, University of Oslo), *Computable error estimates for finite element approximations of elliptic PDEs with rough stochastic data*

Relevant material: Abstract

**2014/2015 Term 3 (B3.03, 1:00pm)**

**Onno van Gaans**(Mathematical Institute, Leiden University),

*'Stability of dynamical systems perturbed by stochastic state dependent jumps'*

*:*

**Francois-Xavier Briol**(OxWaSP, U of Warwick)

*, 'Probabilistic Integration with Theoretical Guarantees'.*

**2014/2015 Term 2 (B1.01, 1:00pm)**

*Learning in Inverse Problems*', UCL

Jan 12th:

**Yulong Lu**,

*Surface evolution equation by level set method*.

**Andrew Papanicolaou**,

*Perturbation Analysis for Investment Portfolios Under Partial Information with Expert Opinions.*

**Hendrik Weber**,

*Non-linear stochastic PDE from statistical physics.*

**MiRaW day**'Uncertainty in complex computer models'.

**Martin Redmann**

*, Balancing Related Model Order Reduction Applied to SPDEs.*

**Murray Pollock**

*, The SCALE algorithm.*

**Matt Dunlop**,

*MAP estimators for Bayesian inverse problems.*

**Andrew Stuart**,

*Overview of EIT: Forward and Inverse Problems.*

**Helen Ogden**,

*Approximating the normalizing constant for graphical models*

*.*

**2014/2015 Term 1 (B1.01, 1:00pm)**

**Andrew Webb and Veronica (Ronni) Bowman**(DSTL),

*Bayesian Inference for Source Term Estimation.*

**Sergios Agapiou**(UoW),

*Preconditioning the prior to delay saturation in Bayesian inverse problems*.

**Aretha Teckentrup**(UoW),

*Introduction to Gaussian Process regression.*

**Shiwei Lan**(UoW),

*Efficient HMC/RHMC algorithms using GP emulation and applications to BIP*.

**Claudia Schillings**(UoW),

*Scaling Limits in Computational Bayesian Inversion.*

*Relevant material: https://www.sam.math.ethz.ch/sam_reports/index.php?id=2014-26p*

**Daniel Simpson**(UoW),

*Prior choice for Gaussian random field models*.

**MiRaW day**'Reduced Order Models and Applications in Uncertainty Quantification'

**Joris Bierkens**(UoW),

*Gaussian importance sampling and pre-conditioning in infinite dimensions*.

**Patrick R Conrad**(UoW),

*Accelerating Asymptotically Exact MCMC for Computationally Intensive Models via Local Approximations*.

**2013/2014 Term 3 (D1.07, 1:00pm)**

**Reading-Warwick Data Assimilation Meeting.**

**Tessy Papavasiliou**(UoW),

*Statistics and Rough Paths*.

**Tim Sullivan**(UoW),

*Polynomial Chaos Methods for Uncertainty Quantification (I)*.

**Tim Sullivan**(UoW),

*Polynomial Chaos Methods for Uncertainty Quantification (II)*.

**Axel Finke**(UoW),

*Optimisation via Pseudo-Marginal Monte Carlo Methods.*

**2013/2014 Term 2 (B1.01, 1:00pm)**

**Matt Thorpe**(UoW),

*Consistency of the k-Means Method Using Gamma-Convergence.*

**Anthony Lee and Daniel Sanz-Alonso**(UoW),

*ABC methods and further discussion*.

Relevant material: http://arxiv.org/pdf/1102.4432v4.pdf.

**Kasia Wolny**(UoW),

*Spectral Gaps for a Metropolis-Hastings Algorithm in Infinite Dimensions*.

Relevant material: http://arxiv.org/abs/1112.1392.

**Michael Betancourt**(UoW),

*The Geometry of Hamiltonian Monte Carlo.*

Relevant material: http://arxiv.org/abs/1112.4118.

**Simon Byrne**(University College London),

*Geodesic Monte Carlo on Embedded Manifolds.*

*http://arxiv.org/abs/1301.6064 (with discussion in http://arxiv.org/abs/1311.1129).*

**Wenlian Lu**(UoW / Fudan University, China),

*From Neuronal Network Model to Neuroimaging: an integration analysis in computational neuroscience.*

**Daniel Sanz-Alonso**(UoW),

*On filtering partially observed, deterministic dynamics*.

**MIR@W Day,**

*Distances between probability distributions.*

**Chenlei Leng**(UoW),

*Dimension reduction using the reproducing kernel Hilbert space.*

**Mark Girolami**(UoW),

*A Probabilistic Solver for Differential Equations?*

**2013/2014 Term 1 (B1.01, 1:00pm)**

**Felipe Medina Aguayo**(UoW),

*The pseudo-marginal approach for efficient Monte Carlo computations.*

*.*

**Henry Abarbanel**(UC San Diego),

*Recent research in Data Assimilation.*

**Andrew Stuart**(UoW),

*Introduction to filtering algorithms: 3DVAR, ExKF and EnKF*.

**Gareth Roberts**(UoW),

*Gradient-based MCMC Algorithms.*

**Murray Pollock and Sergios Agapiou**(UoW),

*Bayesian calibration of computer models.*

**Abhishek Shukla**(UoW),

*Robbins-Monro algorithm for the optimization of functionals*

*defined via expectations*.

**Hendrik Weber**(UoW),

*Kullback-Leibler Approximation for Probability Measures on Infinite Dimensional Spaces*.

**Murray Pollock**(UoW),

*Monte Carlo methods for exactly simulating diffusion and jump diffusion sample paths*.

Relevant material: http://arxiv.org/abs/1302.6964.