Skip to content

Seminar

Jun Yang edited this page Nov 18, 2021 · 163 revisions

Computational Statistics and Machine Learning Group

Join our mailing list by sending a blank email to oxcsml-public+subscribe@googlegroups.com.

Talks for Michaelmas Term 2021

Date Time Location Speaker Topic
15 Oct 2021 4:00 pm small lecture theatre and Zoom Seth Flaxman Statistical challenges from a pandemic
22 Oct 2021 (OxWaSP annual workshop)
29 Oct 2021 (cancelled)
5 Nov 2021 4:00 pm large lecture theatre and Zoom Matteo Giordano Consistent Bayesian nonparametric inference in inverse problems
12 Nov 2021 4:00 pm large lecture theatre and Zoom Jakob Foerster Off-Belief Learning and Zero-Shot Coordination
19 Nov 2021 4:00 pm large lecture theatre and Zoom Kamelia Daudel Infinite-dimensional Alpha-divergence minimisation for Variational Inference
26 Nov 2021 4:00 pm large lecture theatre and Zoom Wenkai Xu Kernel Stein Discrepancies for Goodness-of-fit Testing
3 Dec 2021 (Corcoran Memorial Prize Award and Lecture)
10 Dec 2021 4:00 pm Zoom Jeffrey Negrea Statistical Inference with Stochastic Gradient Algorithms

Previous talks

Talks for Trinity Term 2021

Date Time Location Speaker Topic
23 April 2021 3:30 pm Zoom Roger Grosse Self-tuning networks: Amortizing the hypergradient computation for hyperparameter optimization
30 April 2021 3:30 pm Zoom Julyan Arbel Approximate Bayesian computation with surrogate posteriors
7 May 2021 3:30 pm Zoom Aki Vehtari Practical pre-asymptotic diagnostic of Monte Carlo estimates in Bayesian inference and machine learning
14 May 2021 3:30 pm Zoom Cynthia Rudin Do simpler models exists and how can we find them?
21 May 2021 3:30 pm Zoom James Johndrow Spectral gaps and error estimates for infinite-dimensional Metropolis-Hastings with non-Gaussian priors
28 May 2021 3:30 pm Zoom Alexandra Carpentier Several structured thresholding bandit problems
4 June 2021 3:30 pm Zoom Qiang Liu Recent Applications of Stein's Method in Machine Learning
11 June 2021 3:30 pm Zoom (cancelled)
18 June 2021 3:30 pm Zoom Susan Murphy (Distinguished speaker series)
25 June 2021 3:30 pm Zoom Quan Zhou Complexity of local MCMC methods for high-dimensional model selection
2 July 2021 3:30 pm Zoom Caroline Uhler Causality and Autoencoders in the Light of Drug Repurposing for COVID-19

Talks for Hilary Term 2021

Date Time Location Speaker Topic
15 January 2021 3:30 pm Zoom Lénaïc Chizat Faster Wasserstein distance estimation with the Sinkhorn divergence
22 January 2021 11:00 am Zoom Chris Gamble DeepMind: AI as a Science and AI for Science
29 January 2021 3:30 pm Zoom Victor Veitch Causal Estimation with Machine Learning without (Simple) Unconfoundedness
5 February 2021 3:30 pm Zoom Sinnead Williamsom Bayesian nonparametric models for interaction networks
12 February 2021 3:30 pm Zoom Veronika Rockova METROPOLIS-HASTINGS VIA CLASSIFICATION
19 February 2021 3:30 pm Zoom François-Xavier Briol Kernel-based robust inference for intractable likelihood models
26 February 2021 3:30 pm Zoom Karolina Dziugaite Distribution-dependent generalization bounds for noisy, iterative learning algorithms
5 March 2021 3:30 pm Zoom Linda Siew Li Tan Use of model reparametrization to improve variational Bayes
12 March 2021 3:30 pm Zoom Murat A. Erdogdu Convergence of Online SGD under Infinite Noise Variance, and Non-convexity
19 March 2021 3:30 pm Zoom Maire Florian Approximate asymptotic variance ordering for MCMC
26 March 2021 3:30 pm Zoom Benjamin Guedj A primer on PAC-Bayesian learning followed by News from the PAC-Bayes frontline

Talks for Hilary Term 2020

Date Time Location Speaker Topic
24 January 2020 3:30 pm small lecture theatre Alessandro Barp Sampling Distributions on Manifolds with Hamiltonian Monte Carlo
13 February 2020 TBA TBA Nicolas Heess TBA
14 February 2020 3:30 pm small lecture theatre Michael Arbel TBA
28 February 2020 3:30 pm small lecture theatre Danielle Belgrave TBA
6 March 2020 3:30 pm small lecture theatre David Pfau TBA
13 March 2020 3:30 pm small lecture theatre Benjamin Guedj (cancelled) A primer on PAC-Bayesian learning, and application to deep neural networks

Talks for Michaelmas Term 2019

Date Time Location Speaker Topic
20 September 2019 3:30 pm large lecture theatre Masashi Sugiyama Machine Learning from Weak Supervision: Towards Accurate Classification with Low Labeling Costs
7 October 2019 3:30 pm small lecture theatre Maurizio Filippone Walsh-Hadamard Variational Inference for Bayesian Deep Learning
11 October 2019 3:30 pm Lecture Theatre B, Wolfson Building, 15 Parks Rd David Barber Spread Divergences
18 October 2019 3:30 pm small lecture theatre Andrew Saxe Dynamics of generalization error in overparametrized neural networks
24 October 2019 3:30 pm Lecture Theatre B, Wolfson Building, 15 Parks Rd Bert Kappen Training a quantum system to represent classical data
25 October 2019 11:00 am small lecture theatre Jonathan Huggins Using bagged posteriors for robust inference and model criticism
25 October 2019 3:30 pm small lecture theatre Martin Wiegel Population simulation methods for parallel computing
1 November 2019 3:30 pm small lecture theatre Dino Oglic On Scalable Learning in Krein Spaces and Cyclic Discovery Processes
15 November 2019 3:30 pm small lecture theatre Irina Higgins What is disentangling and does intelligence need it?
29 November 2019 11:00 am large lecture theatre Takeru Matsuda Estimation and selection of non-normalized models
6 December 2019 3:30 pm small lecture theatre Pierre Alquier A Generalization Bound for Online Variational Inference
13 December 2019 3:30 pm small lecture theatre Saifuddin Syed Non-reversible parallel tempering: a scalable, highly parallel MCMC algorithm

Talks for Trinity Term 2019

Date Time Location Speaker Topic
3 May 2019 3:30pm small lecture theatre Xiaowen Dong [Oxford] A maximum-mean-discrepancy goodness-of-fit test for censored data
10 May 2019 3:30pm small lecture theatre Umut Şimşekli [Télécom ParisTech] Nonparametric Generative Modeling via Optimal Transport and Diffusions with Provable Guarantees
17 May 2019 3:30pm small lecture theatre No seminar Departmental Distinguished Seminar
20 May 2019 3:30pm small lecture theatre Cédric Févotte [Toulouse] TBD
24 May 2019 3:30pm small lecture theatre Chris Sherlock [Lancaster] Fast, exact inference for discretely observed Markov jump processes using finite rate matrices
30 May 2019 3:30pm large lecture theatre Ingo Steinwart [Stuttgart] Neural Networks: Initializations and Global Minima
31 May 2019 3:30pm small lecture theatre Eric Nalisnick [Cambridge] Evaluating Deep Generative Models on Out-of-Distribution Inputs
7 June 2019 3:30pm small lecture theatre Kody Law [Manchester] Bayesian Static Parameter Estimation for Partially Observed Diffusions using Multilevel Monte Carlo
14 June 2019 3:30pm small lecture theatre Cancelled due to distinguished seminar TBD
21 June 2019 3:30pm small lecture theatre Chris Nemeth [Lancaster] TBD

Talks for Hilary Term 2019

Date Time Location Speaker Topic
1 February 2019 3:30pm small lecture theatre Tamara Fernandez [UCL] A maximum-mean-discrepancy goodness-of-fit test for censored data
8 February 2019 3:30pm small lecture theatre Olga Isupova [Oxford] TBD
15 February 2019 3:30pm small lecture theatre Lorenzo Rosasco [Genova and MIT] TBD
22 February 2019 3:30pm small lecture theatre Richard Wilkinson [Sheffield] TBD
1 March 2019 3:00pm large lecture theatre Jose-Miguel Hernandez Lobato [Cambridge] Advances in machine learning for molecules
8 March 2019 3:30pm small lecture theatre David Martínez [Oxford] TBD
15 March 2019 3:30pm small lecture theatre Shimon Whiteson [Oxford] TBD

Talks for Michaelmas Term 2018

Date Time Location Speaker Topic
05 October 2018 3:30pm large lecture theatre Song Liu [Bristol] Model Inference with Stein Density Ratio Estimation
19 October 2018 3:30pm small lecture theatre Anthony Caterini / Lawrence Middleton [Oxford] Hamiltonian Variational Auto-Encoder / Parallelising particle MCMC through exact estimation
26 October 2018 3:30pm small lecture theatre Chris Maddison [Oxford] Hamiltonian Descent Methods
2 November 2018 3:30pm small lecture theatre Alexander Lvovsky [Oxford] Quantum optical Ising machines and their simulators for annealing and machine learning
9 November 2018 3:30pm large lecture theatre Aapo Hyvarinen [Gatsby] Nonlinear independent component analysis: A principled framework for unsupervised deep learning
16 November 2018 3:30pm large lecture theatre Brooks Paige [Turing] Semi-interpretable probabilistic models
23 November 2018 3:30pm large lecture theatre Ricardo Silva [UCL] Neural Networks and Graphical Models for Constructing and Fitting Cumulative Distribution Functions
30 November 2018 1:30pm TBD Yi Yu [Bristol] Optimal Change Point Detection and Localization in Sparse Dynamic Networks

Talks for Trinity Term 2018

Date Time Location Speaker Topic
23 April 2018 3:30pm small lecture theatre Aaditya Ramdas [UC Berkeley] Uniform, nonparametric, nonasymptotic confidence sequences (or: how to test if a coin is biased)
27 April 2018 3:30pm small lecture theatre Benedict Leimkuhler [University of Edinburgh] Efficient algorithms for high-dimensional sampling based on extended stochastic dynamics
23 May 2018 11am small lecture theatre Volkan Cevher [EPFL] Mirrored Langevin Dynamics
25 May 2018 3:30pm small lecture theatre Nick Whiteley [Bristol] The Viterbi process and parallelized estimation in high-dimensions
1 June 2018 2:00pm small lecture theatre Quentin Berthet [Cambridge] Optimal Link Prediction with Matrix Logistic Regression
1 June 2018 3:30pm small lecture theatre Stefano Favaro [Torino] A Bayesian approach to disclosure risk assessment
15 June 2018 3:30pm small lecture theatre Ryota Tomioka [MSR] Gaussian Embedding of Knowledge Bases and Application to Question Answering

Talks for Hilary Term 2018

Date Time Location Speaker Topic
26 January 2018 3:30pm small lecture theatre Jovana Mitrovic [Oxford Statistics] Causal Inference via Kernel Deviance Measures
Kaspar Märtens [Oxford Statistics] Augmented Ensemble MCMC with applications to Factorial Hidden Markov Models
2 February 2018 3:30pm small lecture theatre Giuseppe Di Benedetto [Oxford Statistics] Non-exchangeable random partition models for micro clustering
Maxime Rischard [Harvard] Addressing time of measurement bias in records of daily temperature extrema: a spatio-temporal imputation strategy
9 February 2018 3:30pm small lecture theatre Michael Golden [Oxford Statistics] Probabilistic Inference of Nucleotide Coevolution
16 February 2018 3:30pm small lecture theatre Matt Kusner [Warwick and ATI] Grammar Variational Autoencoder
2 March 2018 2pm small lecture theatre Owen Thomas [University of Oslo] Recent Methods in Model-Based Likelihood-Free Inference
2 March 2018 3:30pm small lecture theatre Ingo Steinwart [University of Stuttgart] Learning with (Flexible) Kernels
7 March 2018 3:30pm small lecture theatre John Cunningham [Columbia University] Structure in tensor-variate data: a trivial byproduct of simpler phenomena?
9 March 2018 No CSML seminar. Department Distinguished Seminar.
12 March 2018 3:30pm small lecture theatre Sam Livingstone [UCL] What we talk about when we talk about non-reversible MCMC
16 March 2018 No CSML seminar---room available due to industry event
23 March 2018 3:30pm small lecture theatre Miklos Racz [Princeton University] High-dimensional random geometric graphs

Talks for Michaelmas Term 2017

Date Time Location Speaker Topic
13 October 2017 3:30pm small lecture theatre Catalina Vallejos [UCL and ATI] BASiCS: dealing with technical and biological noise in single-cell expression data
20 October 2017 3:30pm small lecture theatre Yarin Gal [Oxford CS] Bayesian deep learning
27 October 2017 3:30pm small lecture theatre Pawan Kumar [Oxford Engineering Science] Relaxations and Rounding for the Labelling Problem
3 November 2017 3:30pm small lecture theatre Sergio Bacallado [Cambridge] Unbiased estimation and prediction in the hierarchical two-parameter Poisson-Dirichlet process
10 November 2017 3:30pm small lecture theatre Peter Orbanz [Columbia Univ., NYC] Inference from a single, growing graph
17 November 2017 3:30pm small lecture theatre Silvia Chiappa [DeepMind] Explicit-Duration Markov Switching Models
24 November 2017 No CSML seminar. Department Distinguished Seminar.
1 December 2017 3:30pm small lecture theatre Sara Wade [Warwick] Adaptive truncation of a Bayesian nonparametric multivariate regression model for a study of fertility and partnership patterns of Colombian women
8 December 2017 3:30pm small lecture theatre Rajen Shah [Cambridge] The xyz algorithm for fast interaction search in high-dimensional data

Talks for Hilary Term 2017

Date Time Location Speaker Topic
22 March 2017 3:30pm small lecture theatre Kenji Fukumizu [Institute of Statistical Mathematics, Tokyo] Kernel methods for topological data analysis
17 March 2017 3:30pm small lecture theatre Thomas E. Nichols [University of Warwick] Large Scale Evaluation of Random Field Theory Inference in fMRI
16 March 2017 10:00am small lecture theatre Nicholas Chopin [ENSAE] Leave Pima Indians alone: binary regression as a benchmark for Bayesian computation
24 February 2017 3:30pm small lecture theatre Abdul-Lateef Haji-Ali [Oxford Maths] Multilevel and Multi-index Monte Carlo methods for the McKean-Vlasov equation
17 February 2017 3:30pm small lecture theatre Nikolas Kantas [Imperial College London] Towards particle filtering for signals arising from dissipative stochastic PDEs
10 February 2017 3:30pm small lecture theatre Gersende Fort [IMT-CNRS, Toulouse, France] Convergence
of perturbed proximal gradient algorithms
3 February 2017 3:30pm small lecture theatre Dougal Sutherland [Gatsby] Generative Models and Model Criticism via Optimized Maximum Mean Discrepancy

Talks for Michaelmas Term 2016

Date Time Location Speaker Topic
25 November 2016 3:30pm small lecture theatre Chang-Han Rhee [CWI Amsterdam] TBA
18 November 2016 3:30pm small lecture theatre Jean-Michel Marin [Université de Montpellier] Approximate Bayesian Computation using Random Forests
14 November 2016 2:30pm small lecture theatre Mladen Kolar [University of Chicago] Inference in high-dimensional semi-parametric graphical models
11 November 2016 3:30pm small lecture theatre Colin Fox [University of Otaga] MH-MCMC with Stochastic AR(1) Proposals
4 November 2016 3:30pm small lecture theatre Satish Iyengar [University of Pittsburgh] Applications of Diffusion Models to Neuroscience and Finance
28 October 2016 3:30pm small lecture theatre Alexandre Bouchard-Côté [University of British Columbia] Monte Carlo without rejection
21 October 2016 1:30pm-3:30pm small lecture theatre Hong Ge [Cambridge], Simon Byrne [UCL] Turing: A fast imperative probabilistic programming language and Introduction to Julia

Earlier seminars are listed on our old website.