News

Department and friends

Deep Learning Indaba 2017 Summer School (Johannesburg, South Africa)

Summer School 

Dr Konstantina Palla (http://www.stats.ox.ac.uk/~palla/), postdoctoral fellow in the Department of Statistics was invited to give a tutorial on probabilistic reasoning with a focus on its connections to deep neural networks at the Deep Learning Indaba summer school in Johannesburg.

An Indaba is a Zulu word for gathering, meeting for the discussion of affairs of the community. In September 2017, such an Indaba took place in Johannesburg, South Africa under the title "the Deep Learning Indaba". Its target was to attract students from South Africa and Africa in general and build an understanding of the principles and practice of modern machine learning. 

Konstantina Palla - Summer School"I was privileged to give a lecture at the Deep Learning Indaba 2017 summer school (http://www.deeplearningindaba.com). From start to end it was an amazing experience! Machine Learning has been the focus of the global industry lately involving billions of dollars in investment. African Machine Learning has been following this trend by transforming the local industry with various applications. However, the participation of African students and industry representatives in internationally leading machine learning conferences is low. Having spent some years in academia, I can, unfortunately, see that this is the case. Addressing the reasons of this lack of representation is an urgent need. I was happy to meet students talented and eager to learn. And therein lies the problem; the real shortage isn't of talented people. As far as academia is involved, it's the shortage of providing expert education in Machine Learning in Africa. This can be changed and that summer school is a great example. More than 300 attendees came from 23 African countries, as well as other countries across the world, forming one of the largest machine learning teaching events globally. Truly amazed to meet so many students with so much potential. Even more excited to see so many women interested in Machine Learning! A huge congratulations to the organisers. I hope this is one of many to come in South Africa and Africa more generally. We should use our expertise to make a difference! After all,``the best way to predict the future is to create it".

 

Launch of University of Oxford's new Statistical Consultancy unit

Statistical Consultancy 

 

 

The University of Oxford’s Department of Statistics has launched a new consultancy unit, providing comprehensive support and analysis to a range of internal and external businesses. The unit will operate across a wide range of sectors and can offer experience in all aspects of data-based research, allowing businesses and academics access to the University’s world leading expertise in computational statistics, statistical methodology, applied probability, bioinformatics and mathematical genetics.

Oxford University Statistical Consulting (OUSC) covers a large range of different sectors including healthcare, pharmaceuticals, environmental, legal, education, media, public sector, retail and market research and services include access to advice and guidance for statistical analyses, experiment/trial design, data management, data mining, machine learning, qualitative analysis, testing, visualisation, modelling, prediction and reporting.

Although the unit only recently opened its doors for business, its client rosta already includes a host of healthcare and pharmaceutical agents, legal firms and  national media agencies such as the BBC and ITN.

Prof Charlotte Deane, Head of Oxford’s Department of Statistics, said: ‘As Head of Department it is fantastic to see the official launch of Oxford University Statistical Consulting. It is fantastic to have Dr Jennifer Rogers to lead this initiative and we as a department are really excited to be able to offer through our skills this service to help answer the statistical challenges of companies and institutions.’

Commenting on the anticipation surrounding the launch, Dr Jennifer Rogers, Director of OUSC, said: ‘I am really excited to see Oxford University Statistical Consulting getting underway and am looking forward to the challenges ahead. The Department of Statistics here at Oxford has a wealth of knowledge in a wide range of statistical topics and it is great that our clients will be able to tap into this expertise.

‘We live in a world where data is ubiquitous within society and where everything can be measured and monitored. But data is only the raw material of knowledge and it is statistics that allows us to use that data to accumulate evidence and inform decisions. Statistics provides the tools to allow us to forecast the weather, predict footballs scores, monitor crime rates, or decide whether a new treatment is working and our new consultancy unit will provide the skills and expertise needed to tackle these wide ranging and important questions.’

If you would like to contact OUSC you can do so via:

Tel: +44 (0) 1865 281279

e-mail: consultancy@stats.ox.ac.uk

Professor Alison Etheridge awarded OBE

Professor Alison Etheridge

Many congratulations to Professor Alison Etheridge, who has been awarded an OBE in the Queen's Birthday Honours list for services to science. Professor Etheridge, FRS, is  Professor of Probability at the Mathematical Institute and the Department of Statistics and Fellow of Magdalen College.  Her graduate work focused on functional analysis and, from there, she rapidly became interested in the interface between probability and analysis. Much of her work focuses on infinite dimensional stochastic processes and their applications. Most recently her central interest has been a collection of mathematical problems arising in population genetics.


Research reported in New York Times

Jaroslaw Nowak and Professor Charlotte Deane were two of the co-authors of a recent Nature Communications paper on antibody genes and rheumatic heart disease.  The research findings have been widely reported, for example in a New York Times article, and on Australian radio.


Peter Donnelly and Yee Whye Teh  co-authors on RSS Machine Learning report

The Royal Statistical Society's Working Group on Machine Learning has issued a report entitled Machine Learning: the power and promise of computers that learn by example.  Professor Peter Donnelly FMedSci FRS chaired the Working Group, which also included Professor Yee Whye Teh


Sir David Cox receives BBVA Foundation Frontiers of Knowledge Award

The ninth BBVA Foundation Frontiers of Knowledge Award in the Basic Sciences category has been awarded to Sir David Cox and to Professor Bradley Efron of Stanford University, for the development of “pioneering and hugely influential” statistical methods that have proved indispensable for obtaining reliable results in a vast spectrum of disciplines from medicine to astrophysics, genomics or particle physics.

The BBVA Foundation Frontiers of Knowledge Awards seek to recognize and encourage world-class research and artistic creation, prizing contributions of lasting impact for their originality, theoretical significance and ability to push back the frontiers of the known world. These international awards span eight categories: Basic Sciences (Physics, Chemistry, Mathematics), Biomedicine, Ecology and Conservation Biology, Information and Communication Technologies, Economics, Finance and Management, Contemporary Music, Climate Change and Development Cooperation. 

Oliver Kley, Claudia Klüppelberg and Gesine Reinert receive Lloyd’s Science of Risk Prize

Professor Gesine Reinert and her co-authors received the runner-up Lloyd’s Science of Risk Prize 2016 in the category Systems modelling for their paper entitled Risk in a Large Claims Insurance Market with Bipartite Graph Structure. The submission was praised by the judges for providing a valuable contribution to a practical issue facing the insurance industry. They felt that the paper adds directly to insurers’ knowledge, and looks to cover what is a complex, existential issues to most businesses. The need to understand and evaluate the global insurance system is a very difficult problem, and we welcome perspectives on how to address the issue to better understand the interconnected world of risk.


International Prize in Statistics awarded to Sir David Cox

Sir David Cox has been named the inaugural recipient of the International Prize in Statistics, considered the highest honour in its field for major achievements using statistics to advance science, technology and human welfare.

The International Prize in Statistics Foundation is recognizing him specifically for his 1972 paper in which he developed the proportional hazards model that today bears his name. The Cox Model is widely used in the analysis of survival data and enables researchers to more easily identify the risks of specific factors for mortality or other survival outcomes among groups of patients with disparate characteristics.

Successful application of the Cox Model has led to life-changing breakthroughs with far-reaching societal effects, some of which include the following:
* Demonstrating that a major reduction in smoking-related cardiac deaths could be seen within just one year of smoking cessation, not 10 or more years as previously thought
* Showing the mortality effects of particulate air pollution, a finding that has changed both industrial practices and air quality regulations worldwide
* Identifying risk factors of coronary artery disease and analyzing treatments for lung cancer, cystic fibrosis, obesity, sleep apnea and septic shock.

His mark on research is so great that his 1972 paper is one of the three most-cited papers in statistics and ranked 16th in Nature’s list of the top 100 most-cited papers of all time for all fields.




Highly Cited Article

Professor Jonathan Marchini's article ‘A Flexible and Accurate Genotype Imputation Method for the Next Generation of Genome-Wide Association Studies’ It is amongst the top ten downloaded PLOS Genetics articles of all time which has been cited 801 times and downloaded from the site over 7,500 times www.ploscollections.org/article/browse/issue/info:doi/10.1371issue.pcol.v04.i05




 

News Archives