Week 1 - Fridays@4
Speaker: Akshat Mugdal
Title: North Meets South
Speaker: Renee Hoekzema
Title: Exploring the space of genes in single cell transcriptomics datasets
Week 2 - Statistics Student Social
Week 3 - Picnic in the Partks
A special event for Mental Health Awareness Week
Week 4 - Oxford Maths & Stats Colloqium
Speaker: Yurii Nesterov, Universite catholique de louvain
Title: New perspectives for higher-order methods in convex optimisation
Week 5 - Three Minute Thesis
Students will talk for three minutes on their thesis topics.
Week 6 - First Year Graduate Student Poster Session
The first years will present posters of their research.
Week 7 - Maths Meets Stats
Speaker: Melanie Weber
Title: Geometric Methods for Machine Learning and Optimization
Speaker: Francesca Panero
Title: A general overview of the different projects explored during my DPhil in Statistics.
Week 8 - Joint Stats/Maths/CompSci Careers Event
Week 1 - Graduate Lecture
Speaker: Professor Garrett Morris, Department of Statistics
Title: Anti-Bias and Beyond
Week 4 - Distinguished Speaker Seminar
Speaker: Professor Denise Lievesley, Honorary Fellow of Green Templeton College, University of Oxford
Title: Ethics from the perspective of an applied statistician
Week 5 - Graduate Lecture
Speaker: Maria Christodoulou and Mariagrazia Zottoli, Department of Statistics
Title: A Day in the Life of a Statistics Consultant
Week 6 - Joint Statistics/Comp Sci/BDI talk
Speaker: Samir Bhatt, Professor of Machine Learning and Public Health, University of Copenhagen and Professor of Statistics and Public Health, Imperial College London
Title: Modelling infectious diseases: what can branching processes tell us?
Week 7 - Florence Nightingale Lecture
Speaker: Professor Sir Bernard Silverman, Professor of Modern Slavery Statistics, University of Nottingham and Emeritus Professor of Statistics, University of Oxford
Title: Statistics and the fight against modern slavery
Week 8 - Second Year Graduate Student Poster Session
The second years will present posters of their research.
Week 1 - Graduate Lecture
Speaker: Professor Charlotte Deane and Professor Gesine Reinert, Department of Statistics, University of Oxford
Title: How to give and how to listen to presentations
Week 3 - David Blackwell Lecture
Speaker: John Jumper, Senior Research Scientist, DeepMind
Week 4 - Graduate Lecture (Should statistical practice be ethical?)
Speaker: Max van Kleek, Department of Computer Science, University of Oxford
Title: Victims of Algorithmic Violence: An Introduction to AI Ethics and Human-AI Interaction
Week 5 - Graduate Lecture (Should statistical practice be ethical?)
Speaker: Katherine Fletcher
Title: The practicalities of academic research ethics - how to get things done
Week 6 - Graduate Lecture (Should statistical practice be ethical?)
Speaker: Professor David Steinsaltz, Department of Statistics, University of Oxford
Title: Statistics, ethical and unethical: Some historical vignettes
Week 7 - Grauate Lecture
Speaker: Olly Crook
Week 8 - Corcoran Memorial Prize Award and Lecture
Speaker: Professor David Silver
Prize winner: Dr Chris J Maddison
Week 1 - Distinguished Speaker Seminar
Speaker: Sara Van de Geer, Professor at the Seminar for Statistics, ETH Zurich
Title: On classification with small Bayes error and the max-margin classifier
Week 2 - Graduate Lecture
Speaker: Ben Lambert, Department of Computer Science, University of Oxford
Title: Introduction to Bayesian inference for Differential Equation Models Using PINTS
Week 4 - Third year lightning talks
The third years will give talks about their research.
Week 5 - Second year poster session
The second years will present posters of their research.
Week 6 - Graduate Lecture
Speaker: Dr Fergus Boyles, Department of Statistics
Title: Machine Learning in Drug Discovery
Week 7 - Careers Event
Speakers: Careers Service, Keynote speakers and career panellists
Week 8 - First year poster session
The first years will present posters of their research.
Week 1 - Corcoran Memorial Lecture
Speaker: Kerrie Mengersen, Distinguished Professor of Statistics at Queensland University of Technology in the Science and Engineering Faculty
Title: 'Not' Aggregating Data
Week 4 - Graduate Lecture
Speaker: Dr Daniel Nissley, Department of Statistics, University of Oxford
Week 5 - Distinguished Speaker Seminar
Speaker: Professor Bin Yu, Departments of Statistics and Electrical Engineering & Computer Science at UC Berkeley
Title: Veridical Data Science for Biomedical Research: detecting epistatic interactions via epiTree
Week 6 - Department Seminar
Speaker: Dr Davina Durgana, award-winning international human rights statistician and professor with almost 15 years of experience developing leading global models to assess risk to modern slavery
Title: Finding Today's Slaves: Lessons Learned From Over A Decade of Measurement in Modern Slavery
Week 7 - Third year lighting talks
The third years will present their research in short talks.
Week 8 - Second year graduate poster session
The second years will present their research on posters.
Week 1 - Graduate Lecture
Speaker: Charlotte Deane and Gesine Reinert
Title: Active and Passive Presentations: how to present your work, and how to get the most out of seminars, lectures and poster sessions
Week 2 - Black History Month Lecture
Speaker: Jason Forrest, McKinsey & Co, New York
Title: Exploring the Data Visualizations of W.E.B. Du Bois
Abstract: At the 1900 Paris Exposition, an all African-American team lead by scholar and activist W.E.B. Du Bois sought to challenge and recontextualize the perception of African-Americans at the dawn of the 20th-century. In less than 5 months, his team conducted sociological research and hand-made more than 60 large data visualization posters for a massive European audience which ultimately awarded Du Bois a gold medal for his efforts. While relatively obscure until recently, the ramification of his landmark work remains challenging and especially important in light of the Black Lives Matter movement.
Week 5 - Graduate Lecture
Speaker: Jonny Brooks-Bartlett, Senior machine learning engineer, Spotify
Title: Looking back on 4 years in data science
Abstract: It's been almost 4 years since I left academia to work as a data scientist in industry. In that time I've worked in several teams at a couple of companies. I've also spoken to many other data scientists and consulted literature to get a better picture of the current landscape. In this presentation I take you on my journey from the point at which I decided to become a data scientist to now becoming a senior machine learning engineer at a global music streaming service, Spotify. I'll describe the projects I've worked on and do a bit of a deep dive into a ranking system that I built whilst working at Deliveroo. Finally I'll discuss some observations that I have about data science in general that I hope will give a better idea about how data science works in industry and how it differs from what one might do in an academic setting.
Week 6 - Third Year Lightning Talks
The third years talk about their research.
Week 7 - Graduate Lecture
Speaker: Dr. Ekaterina Volkova-Volkmar, Senior Data Scientist, pRED Informatics - Data Science, Roche Pharma Research and Early Development, Roche, Basel, Switzerland
Title: Introduction to Deep Learning and Graph Neural Networks in Biomedicine
Week 8 - Florence Nightingale Lecture
Speaker: Deborah Ashby, President of the RSS
Title: Florence Nightingale and the politicians’ pigeon holes: using data for the good of society
Abstract: Florence Nightingale, best known as the Lady with the Lamp, is recognised as a pioneering and passionate statistician. She was also passionate about education, having argued successfully with her parents to be allowed to study mathematics, and later nursing, herself. More widely, she offered opinions on the education of children, soldiers, army doctors, and nurses, as well as railing against the ‘enforced idleness’ of women. A particular concern was the lack of statistical literacy among politicians. As we celebrate the bicentenary of her birth, the need for education in statistical and data skills shows no signs of abating. What advice would Florence Nightingale offer were she here today?
Week 1 - Graduate Lecture
Speaker: Dominic Yeo, Department of Statistics
Title: An Introduction to Random Graphs
Week 2 - Third Year Lightning Talks
The third year graduate research students talk about their research.
Week 3 - Graduate Lecture
Speaker: Anja Sturm, University of Göttingen
Title: Interacting particle systems: From local stochastic interactions to global phenomena
Abstract: Stochastic interacting particle systems describe the time evolution of particle systems that are distributed on a discrete space (such as the integer lattice). The random changes to these systems are local, and one of the main points of interest is to understand the global and long-time behavior that arises from a particular set of local interaction rules. In this talk, we will introduce some classical interacting particle systems including the contact and the voter model. We will relate them to non-spatial counterparts, in particular Galton-Watson branching processes and Cannings models, and discuss properties such as long-time survival and the existence of (nontrivial) invariant laws. We will then consider these properties for a particular one dimensional model called the cooperative branching coalescent. Here, particles perform independent random walks, only pairs of particles occupying neighboring sites can produce new particles (cooperative branching) and particles that land on an occupied site merge with the particle present on that site (coalescence). We show that this system undergoes a phase transition as the branching rate is increased regarding its survival and upper invariant laws. These results were obtained in joint work with Jan Swart (UTIA Prag).
Week 4 - Graduate Lecture
Speaker: Graham Lee, Research Software Engineer, Oxford RSE Group
Title: Developing better code with automated testing
Abstract: If we want reliable, reproducible simulations and data analysis software, we need to know that we have implemented our code correctly. Further, we need to be confident that changes we make to the code do not introduce unintended flaws. Automated testing is a technique widely used in industry to capture information about the expected behaviour of software and ensure that the system retains that behaviour through its evolution. In this talk, Graham explores the application of the technique to scientific software.
Week 5 - Distinguished Speaker Seminar
Speaker: Nick Jewell, University of California, Berkeley School of Public Health
Title: Cluster-Randomised Test Negative Designs: Inference and Application to Vector Trials to Eliminate Dengue
Week 6 - Graduate Lecture
Speaker: Cora Mezger, Department of Statistics
Title: Opportunities for using machine learning in official statistics
Week 7 - First Year Graduate Poster Session
The first years will present posters of their research.
Week 8 - Collaborative Careers Event with Medical Sciences
1 - Distinguished Speaker Seminar
Speaker: Sarah C. Darby, Professor of Medical Statistics, Nuffield Department of Population Health, University of Oxford
Title: Twenty five year risks of breast cancer mortality in 500,000 women
2 - Corcoran Memorial Prize and Lecture
Speaker: Professor Frank den Hollander, Leiden University
Title: Synchronisation with noise
Followed by a short talk from the Corcoran Prize Winner, Dr Sarah Penington
3 - Graduate Lecture
Speaker: David Kell, CTO, Gyana Ltd
Title: Making the Transition from Academia to Startup
4 - Machine Learning Talk
Speaker: Nicolas Hees and Leonard Hasenclever
Title: Deep reinforcement learning for control - algorithms and architectures
5 - Graduate Lecture
Speaker: Jotun Hein, Department of Statistics
Title: Statistical Biological Sequence Analysis
6 - Machine Learning Talk
Speaker: Arthur Gretton
Title: Some results and conjectures for GAN critic design
7 - Machine Learning Talk
Speaker: Razvan Pascanu
Title: Optimization, learning and data-efficiency
Abstract: Throughout this lecture I would like to ask the question of how we can improve training efficiency for neural networks. This question had been asked many times, from many angles. The angle we will adopt here is to look at the specific role that the learning algorithm or optimizer plays in determining training efficiency.
The lecture will start with a brief overview of learning techniques, going from stochastic gradient descent to second order methods and natural gradient. Each method will be introduced and discussed, highlighting both its strengths and weaknesses.
We will follow, afterwards, with a discussion on a more recent paradigm for efficient learning, namely meta-learning. The topic has seen an extensive growth in the last few years and we will not extensively cover all different forms of meta-learning, but rather focus on those that explicitly learn or parameterize an optimizer. More specifically, we will present gradient based methods such as meta-gradients and meta-curvature, Warp Gradient Descent, and compare them with other families of meta-learning algorithms.
For simplicity, a large part of the discussion will be carried out in the context of supervised learning. However we will also discuss the RL setting, particularly for meta-learning. We will describe the unique challenges imposed by RL and how these could affect the presented algorithms. Finally we will draw some conclusions regarding the role of the optimizer in improving efficiency of learning and discuss potential directions for moving forward.
8 - Distinguished Speaker Seminar
Speaker: Max Welling
Title: Neural Augmentation
9 - Machine Learning Talk
Speaker: Silvia Chiappa
10 - Second year graduate poster session
1 - Active and Passive Presentations: how to present your work, and how to get the most out of seminars, lectures and poster sessions
Speaker: Charlotte Deane and Gesine Reinert, Department of Statistics
2 - How to learn human values and solve all ethical problems - adequately
Speaker: Stuart Armstrong, Future of Humanity Institute, University of Oxford
3 - How to have a successful internship during your DPhil
Speaker: Clare West, Susan Leung and Lyuba Bozhilova
4 - A few examples of how statistics is used when phasing and imputing with sequencing reads in genetics
Speaker: Robbie Davies, Department of Statistics
5 - Student Entrepreneurs
Speaker: Cath Spence, Principal Licensing & Ventures Manager - Incubator Lead
6 - Hidden relatedness, natural selection and disease heritability in the human genome
Speaker: Pier Palamara, Department of Statistics
7 - Distinguished Speaker Seminar
Speakers: Professor Tyler VanderWeele, Departments of Epidemiology and Biostatistics at the Harvard T.H. Chan School of Public Health and George Eastman Visiting Professor at Balliol College, University of Oxford
Title: Sensitivity Analysis in Observational Research: Introducing the E-Value
8 - Third year talks
The third years will give short talks on their research.
1 - Florence Nightingale Lecture
Speaker: Sir David Spiegelhalter, OBE, FRS
Title: What would Florence Nightingale make of the way data is being used today?
2 - Graduate Lecture
Mental Health talk with panel discussion
3 - Distinguished Speaker Seminar
Speaker: Professor Donald Rubin, Harvard University
Title: Essential Concepts of Causal Inference: A remarkable history and an intriguing future
4 - Strengthening Analytical Thinking for Observational Studies: the STRATOS initiative – why is it needed, what does it do & who can be involved? (Ethics and Statistics Series)
Speaker: James Carpenter (Department of Medical Statistics, LSHTM)
Abstract: While statistical methodology for both the design and analysis of observational studies has seen major advances over the last 20 years, many of these methodological developments are ignored in applications. Specifically, the lack of guidance on vital practical issues discourages many applied researchers from using more sophisticated and possibly more appropriate methods when analyzing observational studies. Furthermore, many analyses are conducted by researchers with a relatively weak statistical background and limited experience in using statistical methodology and software. Consequently, even ‘standard’ analyses reported in the medical literature are often flawed, casting doubt on their results and conclusions.
An efficient way to help researchers to keep up with recent methodological developments is to develop guidance documents that are spread to the research community at large. These observations led to the initiation of the strengthening analytical thinking for observational studies (STRATOS) initiative, a large collaboration of experts in many different areas of biostatistical research.
The objective of STRATOS is to provide accessible and accurate guidance in the design and analysis of observational studies. The guidance is intended for applied statisticians and other data analysts with varying levels of statistical education, experience and interests. In this talk I will outline the background to the initiative, reflect on challenges and achievements to date, and outline how to become involved.
5 - Collaborative Careers Event
A careers event held jointly with Maths and Computer Science.
6 - Distinguished Speaker Seminar
Speaker: Professor Mike Steel, Director of the Biomathematics Research Centre at University of Canterbury, Christchurch, New Zealand
Title: From evolutionary trees to networks and back again
Speaker: Professor David Aldous, University of California, Berkeley
Title: Scattered thoughts from applied probability: networks, security queues and prediction tournaments
8 - Graduate poster session
1 - Multiple phenotype models and the genetic basis of brain Structure and function
Speaker: Kevin Sharp, Department of Statistics
Abstract: Much is unknown about the genetic basis of brain structure and function. Magnetic Resonance Imaging (MRI) provides a powerful, non-invasive technique for obtaining estimates of relevant quantities such the volumes, and thicknesses of different structures within the brain or the integrity of the connections between them. Recently, we carried out genetic association studies of 3,144 such structural and functional brain imaging phenotypes in the first tranche of 8428 imaged subjects from the UK Biobank [1]. In terms of the number of phenotypes analysed, this is by far the largest study of its kind so far undertaken.
In this talk, I will discuss this work together with our ongoing analysis based on twice as many individuals. In particular, I will focus on the utility of multivariate models. Such approaches fit a joint model for the association between a genetic variant and a group of correlated traits and can utilise estimates of genetic correlation to boost power. If time permits, I will also describe a sparse extension of this model which can be more powerful when only a small subset of phenotypes is truly associated. I will try to provide sufficient background so that no prior knowledge of genetics is required! [1] Lloyd T. Elliott, Kevin Sharp, Fidel Alfaro-Almagro, Sinan Shi, Karla Miller, Gwenaëlle Douaud, Jonathan Marchini, Stephen Smith. Genome-wide association studies of brain imaging phenotypes in UK Biobank. Nature volume 562, pages 210–216 (2018).
2 - Util - using machine learning to help investors make a positive impact on people, planet and pocket
Speaker: Abdel Turkmani and Stephen Barnett, Util
Abstract: Util is a London based Financial Technology startup, founded in Oxford in 2017 which seeks to use data science and machine learning to better understand the positive and negative impacts of their investments. Abdel Turkmani and Stephen Barnett, Util’s cofounders, will be talking about their startup journey, providing an insight into how state of the art data science is being incorporated into a fintech business model and presenting a day in the life of an engineer at Util. Both founders will be joining attendees after the talk to discuss a startup career path.
3 - Exchangeability
Speaker: Christina Goldschmidt, Department of Statistics, University of Oxford
Abstract: Exchangeability is a fundamental concept in probability and statistics, and crops up in all sorts of different places. In this lecture, I will give an introduction to exchangeability, and prove a form of de Finetti's theorem, which relates exchangeable sequences of random variables to i.i.d. sequences of random variables. I will finish with some ideas for where you can find out more if you need to.
4 - Robust Research - Why do we care?
Speaker: Verena Heise, Nuffield Department of Population Health
5 - Fair machine learning: how can data-driven decision-making perpetuate bias, and what should be done about it? (Ethics and Statistics Series)
Speaker: Reuben Binns, Department of Computer Science
Abstract: Statistical models are useful for guiding decisions that involve prediction or classification, and they often outperform human decision makers. Machine learning promises to improve such models even further by uncovering the hidden relevance of factors that we might not consider. But the data used to train such models can be biased - either by failing to represent the relevant population, or by reflecting unjust social structures - which may lead to unfair decisions. This talk will introduce a growing body of interdisciplinary work at the intersection of computer science, law, and statistics, which investigates these issues and ways to address them. Reading: Chapter 1 (and optionally, chapter 2) of the 'Fair Machine Learning' book https://fairmlbook.org/pdf/fairmlbook.pdf
6 - Second Year Poster Session
7 - Data Science in Industry: Innovation through Impact
Speaker: Blenheim Chalcot
Abstract: Data Scientists implement projects on the intersection between academic research and delivering business value. The variety, intellectual stimulation and impact that a Data Science role can provide makes it an engaging and fulfilling role to be in. This talk covers some of the Data Science projects carried out at BC and summarises which Machine Learning methods we are currently looking into: - Using Natural Language Processing to speed up clinical trials; - Using cosine similarity methods to find and sign new music talent; - Using Markov Chains and Causal Models to save small to medium businesses from overspending on marketing with Multi Touch Attribution and Marketing Mix Modelling. We used to be students in a similar position to you, with a choice between further academia or a job. We decided to choose both! Listen to anecdotes from our experiences in BC Data Science and how these projects are impacting society and research.
8 - Robust Research - Why do we care? (Ethics and Statistics Series)
Speaker: Verena Heise (Nuffield Department of Population Health, University of Oxford)
Abstract: Are most published research findings false? Why should we care? And is there anything we can do about it? In this talk I will argue that open science and good research practices can help make our research findings more robust. While there are a number of solutions that can be implemented by individual researchers, there are wider issues, for example around incentives and skills training, that require cultural change. To lobby for this change we have started Reproducible Research Oxford, an initiative that is mostly driven by early career researchers.
I will give an overview of our current and planned activities that aim to drive cultural change to improve robustness of research findings. Background papers: A manifesto for reproducible science (essential) Why Most Published Research Findings Are False (further reading).
1 - Active and Passive Presentations: how to present your work, and how to get the most out of seminars, lectures and poster sessions
Speakers: Professor Charlotte Deane and Professor Gesine Reinert, Department of Statistics, University of Oxford
2 - What should you eat to be healthier?
Speaker: Richard Davies and Christopher Bonnet, Project Sapiens
Abstract: Food matters to your health. Project Sapiens combines science, large scale data and machine learning to tell you what to eat based on your unique metabolism. We are running PREDICT, one of the largest nutrition response studies ever carried out, in collaboration with Tim Spector, Professor of Genetic Epidemology at King’s College and Director of Twins UK, and a multi-disciplinary team of top scientists. PREDICT and future studies will enable us to understand the major differences between individual's food responses, which are affected by a myriad of factors, including gut microbiome. Read more at https://project-sapiens.com/ and https://predict.study/
3 - Disrupting Drug Discovery by automating intuition
Speaker: Willem van Hoorn, Chief Decision Scientist, Exscientia
Abstract: Artificial Intelligence and machine learning are taking over the world. The process of discovering new drugs was thought, by practitioners, not to be amenable to AI disruption. At Exscientia we begged to differ and we are now the first company to have automated drug design. This was not done by throwing large amounts of data at a deep neural network and watch the magic happen. The real bottleneck in drug discovery is the medicinal chemist, who is confronted with both too much data (back-catalogue of 150 years of drug discovery) as well as too little data (few expensive data points are generated in a project). This led practitioners to believe drug discovery is an art where intuition plays a big role. We have built a platform that captures that intuition and shown that the typical drug discovery productivity can be dramatically improved. We are now at the end of traditional medicinal chemistry but not at the end of medicinal chemists, there is still a big role for the human! Although chemical structures will be shown no prior chemistry knowledge required to attend this talk.
4 - Short talks about internships
Speakers: Mikolaj Kasprzak and Xiaoyu Lu
1 - Data Science & Industry
Speakers: Dr Sepanda Pourhyaya and the alumni Alexandra Darmon who both work for the analytics focused startup, Fospha
Abstract: If you've ever wondered how to become a data scientist and build concrete solutions to real world problems in a fast-paced environment, come along and find out about:
- The new challenges of Data Science
- What should you do to be a Data Scientist?
2 - "Sufficientness" postulates for Gibbs-type priors
Speaker: Stefano Favaro, Department of Statistics, University of Oxford
3 - Combinatorics in ChemoInformatics
Speaker: Jotun Hein, Department of Statistics, University of Oxford
4 - Interplay between sampling and optimization
Speaker: Daniel Paulin, Department of Statistics, University of Oxford
Abstract: Suppose that U is a real valued convex function in a d dimensional Euclidean space. The problems of (1) finding the minimum of U (2) sampling from a distribution whose log-likelihood is U (up to an additive normalising constant) look quite different at first sight. However, it turns out that there are important connections between these problems. In this talk, we will review some recent results in the literature exploring these connections, and highlight some open problems.
5 - The dynamics of spin plaquette models
Speaker: Paul Chleboun, Department of Statistics, University of Oxford
1 - Dirichlet process models and their applications
Speaker: Jessie Wu, Department of Statistics, University of Oxford
Abstract: This is an introductory lecture to Dirichlet Processes (DP). The beginning of the lecture will introduce the definition and the various representations of DP to demonstrate how the process can be constructed. Popular variations of DP will also be presented along with the applications of these models, such as clustering and nonparametric regression.
2 - An introduction to exchangeable random partitions and random discrete probability measures
Speaker: Ben Bloem-Reddy, Department of Statistics, University of Oxford
Abstract: Random partition processes underly many constructions for Bayesian models of clustered data; the Chinese Restaurant Process (CRP) is the prototypical example. They are also perhaps the simplest type of combinatorial stochastic process, and are a starting point for studying processes with more complex structure. In the first part of the lecture, I will give an introduction to exchangeable random partitions and their connection to random discrete probability measures via Kingman's paintbox, with special focus on the CRP. In the second part, I will focus on the properties of the Dirichlet Process (DP), and how different constructions of the DP lead to various classes of random probability measures that appear in the probability and Bayesian nonparametrics literature. Some of these properties will be used in the subsequent week's grad lecture by Jessie Wu on applications of and inference with models that use the DP as a building block.
3 - What is new in R?
Speaker: Marco Scutari, Department of Statistics, University of Oxford
Abstract: The capabilities of R have evolved considerably since its first public releases in the 1990s, with each major version bringing in new functionality. In this lecture I will discuss two areas which have seen major evolutions in R 3.x.y: the inclusion in the core distribution of the "parallel" package , which provides facilities for parallel computing; and the release of the Hadleyverse packages, which streamline common data cleaning tasks.
4 - Statistical Alignment with Long Insertion-Deletions for 2 and More Sequences
Speaker: Jotun Hein, Department of Statistics, University of Oxford
5 - Hands-On Statistics: Getting started in Outreach and Public Engagement
Speaker: Mareli Grady, Department of Statistics
Abstract: The benefits of engaging in outreach and public engagement are numerous and the impacts important. In this talk we will explore the opportunities available in the Department and at Oxford in general, show some examples of outreach and public engagement activities currently in use and give you the opportunity to design and discuss some ideas for yourself. Put your creative hat on and join in!
1 - Statistical Consultancy
Speaker: Jen Rogers, Department of Statistics, University of Oxford
Abstract: Jen took on the role of Director of Statistical Consultancy Services within the Department in July last year. In this talk she will be presenting her experiences of the job, talking about what is like to work with industry on a consultancy basis and professional aspects associated with the role. She will go through case studies of work that she has carried out, introducing the kinds of statistical problems that she typically encounters. Finally, she will also be outlining ways in which you may be able to get involved with consultancy activities within the Department.
2 - Causal Inference: a Machine Learning Perspective
Speaker: Dr Ricardo Silva, Lecturer in Statistics, UCL
Abstract: In this exposition, we will discuss the common tools used in the machine learning community to describe causal assumptions and how this leads to particular ways of thinking concerning the estimation of causal effects. We will focus on two main case studies: how to combine data from observational and experimental studies; and how to criticise ways of adjusting for confounding in observational studies, given that background knowledge my be imperfect and hide default assumptions with unintended consequences.
3 - Replacing the human trader
Abstract: XTX Markets trades an average of $80B daily in thousands of financial instruments on an electronic basis with little human interaction. This adds liquidity to the global financial markets and is made possible by taking a highly systematic, purely data-driven, approach to trading. In this talk, we will describe how trading can be automated, the sort of problems that it poses, and how a small team of mathematicians, statisticians and computer scientists tackles them. Graduate Talk by XTX Markets.
4 - The Savage Axioms for Dummies
Speaker: Geoff Nicholls, Department of Statistics, University of Oxford
Abstract: In Bayesian inference the subjective expected utility is an object of foundational importance. The decision maker’s choices maximise this utility, and so it decides some elements of our statistical methodology. For the SEU to exist we need a prior and a utility representing the subjective beliefs of the analyst to exist.
The Savage Axioms are a set of axioms which impose a certain “consistency” on our beliefs and values, and are sufficient for a representative prior and utility function to exist. I will present the axioms in the form given by Maurice de Groot,and look at two famous paradoxes which show, or appear to show, that some very natural human beliefs and values are in conflict with these axioms.I will mention some approaches to resolving of these paradoxes and conclude with a brief evaluation of the axioms.
5 - Active and Passive Presentations: how to present your work, and how to get the most out of seminars, lectures and poster sessions
Speaker: Professor Charlotte Deane and Professor Gesine Reinert, Department of Statistics, University of Oxford
1 - Genetics of obesity and human fat distribution
Speaker: Cecilia Lingren, Nuffield Department of Medicine, University of Oxford
Abstract: Prof. Cecilia Lindgren is a Senior Group Leader at the Big Data Institute, University of Oxford. Her research focuses on applying genomics to dissect the etiology of obesity related traits and their relationship with reproductive health. In this seminar, Prof Lindgren will:
- Introduce the background and rationale behind obesity trait genetic research
- Give an overview of where we are in the field with our most recent research (including unpublished data)
- Discuss what the clinical utility can be, and not be, of these discoveries
- Outline the big data institute and what work she is doing, which is related to obesity genetics
2 - Big Network Modeling and Anomaly Detection for Cyber-Security Applications
Speaker: Patrick Rubin-Delanchy, Department of Statistics, University of Oxford
Abstract: Data arising in cyber-security applications often have a network, or `graph-like', structure, and accurate statistical modelling of connectivity behaviour has important implications, for instance, for network intrusion detection. We present a linear algebraic approach to network modelling, which is massively scalable and also very general. In this approach, nodes are embedded in a finite dimensional latent space, where common statistical, signal-processing and machine-learning methodologies are then available. A central limit theorem provides asymptotic guarantees on the statistical accuracy of the embedding. We explore an intriguing connection between `disassortivity', whereby nodes that are similar are relatively unlikely to connect, and space-time, as defined in special relativity. Mass testing for anomalous edges, correlations, and changepoints is then discussed. Results are illustrated on network flow data collected at Los Alamos National Laboratory.
3 - What's new in Monte Carlo
Speaker: Arnaud Doucet, Department of Statistics, University of Oxford
4 - An introduction to branching random walk models
Speaker: Julien Berestycki, Department of Statistics, University of Oxford
5 - Combinatorics of Recombination
Speaker: Jotun Hein, Department of Statistics, University of Oxford
1 - Introduction to Machine Learning
Speaker: Cedric Archambeau, Amazon
2 - Geometric Embeddings of Biological Data
Speaker: Junhyong Kin, Department of Computer & Information Science, University of Pennsylvania
3 - Stable Lévy processes and forests
Speaker: Christina Goldschmidt, Department of Statistics, University of Oxford
4 - Computationally efficient state space modelling
Speaker: Jouni Helske, University of Jyvaskyla
Abstract: State space modelling (SSMs) offers an unified framework for statistical inference of a broad class of time series and other data. For example, traditional ARIMA models, structural time series models, and generalized linear mixed models can all be represented in a state space form. In this talk I will first introduce the basic concepts of state space modelling, and discuss some potential problems when using SSMs in practice. I then outline some (partial) solutions for these problems and their connection to my own research of computationally efficient Bayesian inference of SSMs.
5 - Statistical/computational phylogenetics
Speaker: Alex Bouchard-Cote, Department of Statistics, The University of British Columbia
Abstract: I will start with a basic introduction to phylogenetics from a statistical and computational point of view. I will then describe some non-standard motivating applications in historical linguistics and cancer genetics. Finally, I will cover a methodology motivated by challenges encountered in these applications, which we call Divide-and-Conquer Sequential Monte Carlo methods. The slides for this talk can be found here.