WiDs speakers

Cecilia Mascolo (Turing Fellow and University of Cambridge, UK)

Cecilia Mascolo is a mother of a teenage daughter. She is also Full Professor of Mobile Systems in the Computer Laboratory, University of Cambridge, UK, a Fellow of Jesus College Cambridge and a Fellow at the Alan Turing Institute for Data Science in London. Prior joining Cambridge in 2008, she was a faculty member in the Department of Computer Science at University College London. She holds a PhD from the University of Bologna. Her research interests are in human mobility modelling, mobile and sensor systems and networking and spatio-temporal data analysis.

She has published in a number of top tier conferences and journals in the area and her investigator experience spans projects funded by Research Councils and industry. She has received numerous best paper awards and in 2016 was listed in “10 Women in Networking /Communications You Should Know”. She sits on the editorial boards of IEEE Pervasive Computing, IEEE Transactions on Mobile Computing, ACM Transactions on Sensor Networks and ACM Transactions on Interactive, Mobile, Wearable and Ubiquitous Technologies.

Talk title:

Harvesting and Making Sense of Mobile and Wearables Data: Challenges and Opportunities

Abstract:

Theoretically, the advent of powerful and inexpensive sensing and wearable technology has enabled
the gathering of very fine grained data about human behaviour at large scale and long periods.
However, both data scientists and application developer are facing huge challenges related with both harvesting data efficiently while preserving user privacy and making sense of such noisy data.

In this talk I will illustrate the challenges and opportunities of this research with respect to a) the efficient and privacy preserving data collection by bringing the logic and learning on device and close to the user and b) the making sense of such data for a variety of applications including mobile health and urban computing.

Codina Cotar (Department of Statistical Science, UK)

Codina Cotar is a Reader in Probability in the Department of Statistical Science, UCL. She completed her PhD in Probability in 2004 in the Statistics Group, the Department of Mathematics, University of Bristol.

Between 2004-2011 she held postdoctoral fellowships at University of British Columbia, Technishe Universitaet Berlin and Technische Universitaet Munich. Between 2011-2012 she held a Research Immersion Fellowship at the Fields Institute, Toronto, Canada. Between 2012-2016 she was a Lecturer in the Department of Statistical Science, UCL.”

Talk title:

Breaking the curse of dimension in quantum mechanical computations through analysis and probability

Abstract:
The chemical behaviour of atoms and molecules is described accurately, at least in principle, by quantum mechanics, which is modelled through Schroedinger’s equation for the many electrons problem. If Schroedinger’s equation could be solved accurately and efficiently then almost any property of the materials could be determined accurately. Unfortunately, there is neither an accurate nor an efficient method to solve these problems, whose computational costs remain high for large systems (e.g. of size of millions of atoms) due to a curse of dimension phenomenon. To simulate chemical behaviour, approximations are needed which is done by means of density functional theory (DFT) – a large and very active research area in physics and chemistry.

I will start my talk with an introduction to the area, then I will present very recent approaches to the area to tackle some of the existing issues. This is based on joint works with Mircea Petrache (Pontificia Universidad Catolica de Chile)

Emma McCoy (Imperial College London, UK)

Emma McCoy is the Vice-Dean (Education) for the Faculty of Natural Sciences and Professor of Statistics in the Mathematics Department at Imperial College London. Her current research interests are in developing time-series and causal inference methodology for robust estimation of treatment effects, particularly in a transport setting. Emma is a member of Council of the Royal Statistical Society as well as a member of their Academic Affairs Advisory Group.

She has a particular interest in Mathematics and Statistics education and is a member of the Royal Society’s Advisory Committee for Mathematics Education (ACME). She is leading the grand challenge associated with monitoring complex systems as part of the Lloyd’s Register Foundation/Alan Turing Institute Data Centric Engineering Programme and has recently been invited to join the Royal Society’s steering group on the dynamics of data science. She is involved in many outreach activities; she has delivered the London Mathematical Society (LMS) popular lecture and is a regular speaker at the Royal Institution’s Mathematics Masterclasses and public engagement of science events.

Talk title:

Causal inference in observational studies

Abstract:
We are all familiar with the idea that correlation does not imply causation. However, knowledge of this fact does not negate the need to address the question of how we approach causal questions when all we have are observational data. I will outline the statistical approach to causal inference in the framework of potential outcomes and counterfactuals and describe recent extensions to propensity score methods with applications in transportation studies.

Fatima Batool

Fatima is a PhD candidate in Statistics at University College London and on enrichment year at the Alan Turing Institute. She is also Lecturer in Statistics at the Department of Statistics at CIIT Lahore Pakistan. She is one of the four recipients of the prestigious Commonwealth scholarships for commencing her PhD studies from Pakistan in 2015. She was awarded the Gold and the Chancellor medal, for securing first positions in MPhil and MSc (by research) respectively, in Statistics by the Quaid-i-Azam University, Islamabad, Pakistan.

She has varied research interests. In her PhD, she is working on developing methods to find patterns in data. She loves to explore data and believes in extracting true knowledge from it without misusing it for one’s personal objectives. She also designs randomized response surveys, to collect data and estimate statistics of interest on sensitive issues.

Kate Saunders (University of Oxford, UK)

Her early research focused on suicidal behaviour, while her doctorate explored psychiatric phenotypes. She conducted one of the first studies to distinguish bipolar disorder and borderline personality disorder on the basis of cognitive function and social behaviour. In 2014 Kate was given a NARSAD Young Investigator award by the Brain and Behaviour Foundation to study whether deficits in social cognition are a valid treatment target in borderline personality disorder.

Alongside this, in conjunction with colleagues at the Institute of Biomedical Engineering she  set-up, and is the clinical lead for, the AMoSS study, which uses smart-phones and wearable devices e.g. actigraphs, to prospectively explore mood and related behavioural variables in a cohort of bipolar and borderline patients. Kate is also involved in a joint ATI-Oxford collaboration to analyse speech data for markers of mood states.

Talk title:

Digital phenotyping & mental disorders

Abstract:
The emergence of mobile technologies enables us to collect high frequency low friction prospective data. The ubiquity of mobile phone networks, and the empowerment associated with mood monitoring means that the adoption of such approaches is well tolerated by patients and uptake is high. This method of data capture combined with the expansion in machine learning technologies present the possibility of not only minimising the inherent bias of retrospective descriptions of psychopathology, but also the chance to identify proxy markers for emergent crises. We have developed a platform for the collection of frequent mood data using smart phones. This prospective mood data has been augmented by behavioural (e.g., activity,) and environmental (e.g., daily stress, contextual threat) data collected remotely using sensors and data from smartphones, including accelerometers, specific ‘apps’, and behavioural data (e.g. keystroke errors, time an app is used).
I will present findings from the AMoSS study: a large longitudinal prospective monitoring study of individuals with bipolar disorder, borderline personality disorder and healthy volunteers.

Kate Rohinshine (Microsoft,UK)

Kate manages the Data and AI Cloud Solution Architect team for the Financial Services and Insurance at Microsoft UK. Prior to joining Microsoft, she worked in companies focused on applying behavioural analytics to augment decision making for both government and finance in the maritime space. Between 2007 -2010 she worked in scientific research in Neurobiology, with an emphasis on post-transcriptional RNA modifications and it’s influence on behaviour. She holds a MSc in Molecular Biology from Bar Ilan University and a MBA from Tel Aviv University.

Kathy Whaler (University of Edinburgh, UK)

Kathy Whaler has been the Professor of Geophysics at the University of Edinburgh since 1994. Her research focuses on the magnetic field of the Earth and other planets, from measurements made at the surface and by low orbiting satellites.  She has been the President of the Royal Astronomical Society (that has solid earth geophysics within its activities) and of the International Association of Geomagnetism and Aeronomy. She is currently Vice-President of the International Union of Geodesy and Geophysics. She was made a Fellow of the Royal Society of Edinburgh in 1997, and of the American Geophysical Union in 2005. She was ‘Gauss Professor’ of the Göttingen Academy of Sciences and gave the Royal Society of Edinburgh Gunning Victoria Prize Lecture, both in 1998, had a minor planet named after her by the International Astronomical Union in 2006, received the Price Medal of the Royal Astronomical Society in 2013, and was made OBE in the latest New Year’s Honours list.

Title:
How big data has transformed and is transforming geophysics

Abstract:
Computing power and new technology have transformed geophysical data acquisition, processing, modelling and interpretation over the last few decades. When I started research in geophysics, most instruments operated in analogue mode, collecting data on rotating drums using ink pens or a nib scratching smoked paper. Data digitisation was done by hand, and the traces were reduced to a few digital points that computers then could handle. Fast forward just four decades and the landscape is transformed – broadband digital data can be collected and recorded by small, low power consumption units. The instruments are installed in permanent observatories, on satellites, and deployed in portable experiments. Observatories have been automated, and observers can interrogate the data remotely. The duration of temporary deployments has increased thanks to computer storage, battery and solar panel improvements, and again the data can be telemetered to an observer or stored for long periods between site visits. Geophysicists can issue alerts – e.g. tsunami or space weather warnings – and make forecasts in near real time. Vast quantities of satellite data can be processed and analysed to determine how the ground has moved following a tectonic event such as an earthquake, even if no ground sensors are present. Satellites now orbit other solar system objects, so geophysicists have ‘taken over’ from astronomers to use their instruments and methods to infer their inner workings. Recent advances in modelling and interpretation take advantage of the increase in computing power, such as producing a range of models satisfying the data, or a probability distribution describing them, proper statistical analysis of the uncertainties, and algorithms and machine learning that search for patterns in or features of the data. I will give examples of some of the above, taken primarily from the research areas of my group, centred on electromagnetic field measurements from the Earth and other solar system objects.

Heather Savory (Data Capability, UK)

Heather joined the Office for National Statistics as Deputy National Statistician and Director General for Data Capability in May 2015. She has a strong commercial track-record and extensive board and senior management experience in entrepreneurial and high technology businesses, consulting and central government.
Heather was appointed as Chair of the Open Data User Group (ODUG), an independent advisory group to the Government’s Public Sector Transparency Board, for three years from May 2012; she was also a member of the Regulatory Board of the Royal Institute of Chartered Surveyors (RICS).
Previously, Heather worked at 3Dlabs as Vice President of Engineering and Operations and eComData as Managing Director. More recently, Heather worked at the centre of UK Government, spending two years in HM Treasury and three in the Department of Business, Innovation and Skills at the Better Regulation Executive (BRE).

Abstract:
The Sustainable Development Goals (SDGs), otherwise known as the Global Goals, are a universal call to action to end poverty, protect the planet and ensure that all people enjoy peace and prosperity by 2030. The goals are universal, for all countries and people, and aim to ensure that we leave no one behind. At the heart of the 2030 Agenda is the need for better data, only by harnessing new data sources and techniques in collaboration with partners across the globe will the challenge be met.

Heather Savory, Deputy National Statistician and Director General for Data Capability at the Office for National Statistics (ONS) will provide insight on how ONS, as the UK’s National Statistics Institute (NSI), is meeting the challenge to report UK progress against the Global Goals by using new techniques and developing new data sources as well as leading a global initiative to develop a platform that will enable NSIs from across the globe to share and access data for the public good.

 

Jil Matheson (UK Statistics Authority, UK)
Jil Matheson served as National Statistician, Head of the Government Statistical Service and Chief Executive of the UK Statistics Authority from 2009 until her retirement in 2014, following a career in social research and statistics. During that time Jil also Chaired the OECD’s Committee on Statistics and Statistical Policy and the UN Statistical Commission.
Jil is a Trustee of NatCen Social Research, and is a member of a High Level Expert Group, chaired by Professor Joe Stiglitz, on Measuring the Progress of Societies, and of the Royal Society’s Advisory Committee on Mathematics Education.

In 2015 she led a group for the British Academy on the need for improved quantitative skills in the UK. And in 2015/16 carried out a BBC Trust impartiality review of the BBC’s use and reporting of statistics, ‘Making Sense of Statistics’ published in August 2016.
Jil is a fellow of the Academy of Social Sciences and the Royal Statistical Society. She was made DCB in 2014.

Mariana Damova (Mozajka, Bulgaria)

Dr. Mariana Damova is the CEO of Mozaika, The Humanizing Technologies Lab, a company providing research and solutions in the field of data science, reasoning with natural language semantics, natural human computer interfaces and human insight. Her background is in natural language processing, Semantic Web Technologies and AI, with strong academic and industrial executive record, having taught graduate courses and conducted research at several universities and successfully lead international interdisciplinary teams with projects carrying technological risks on various facets of intelligent information management in North America and in Europe.

Dr Damova holds a PhD from the University of Stuttgart, and a mini MBA from McGill University. She teaches currently Semantic Web Technologies at the New Bulgarian University and at Sofia State University, regularly reviews books and articles for ACM ComputingReviews.com and has authored books and scientific articles in linguistics and semantic technologies.

Title:

Semantic information infrastructures and their value to data intensive services

Abstract:

Data is world’s greatest natural resource. Data is the world’s great new natural resource. What steam power was to the 18th century, electromagnetism to the 19th and fossil fuels to the 20th… data will be to the 21st.” Ginni Rometty, President and Chief Executive Officer of IBM. Dealing with heterogeneous data sources and extracting value from them is one of the most important challenges in recent years.

From business information delivery, through human resources management, life sciences and Industry 4.0 to earth observation to be able to consume data it is necessary to provide with information infrastructures that allow reliable collection, storage, analytics and visualization of the data – e.g. the realization of the full cycle of the so called data value chain. Semantic technologies start playing crucial role in these endeavours. The present talk will discuss the nature of semantic information infrastructures, their advantages and drawbacks, and will show the trend of their adoption by big industrial companies along with examples of applications based on semantic information infrastructures produced by Mozaika.

Marie-Christine Sawley (Exascale Lab, France)

Marie-Christine Sawley has been the Intel Director of the Exascale Lab in Paris since its opening at the end of 2010. Prior to this, she worked for 3 years as senior scientist for the ETH Zurich in the CMS computing team at CERN. Between 2003 and 2008, she was the director of the CSCS, the Swiss national supercomputing centre, driving the growth and computing capacities of the centre during this period. During her career at EPFL, between 1988 and 2003, she managed a number of high profile HPC projects across a variety of disciplines. Over the last 20 years, she has built expertise as HPC technology manager to serve the advancements of world class science, mission critical computing, extreme sporting competition or industrial partnerships.
Marie-Christine holds a degree in Physic and a PHD in Plasma Physics (1985) from EPFL. She spent two periods in Australia: for a postdoc at Sydney Uni in 1987-88 and for a sabbatical leave in 1999.

Title:
Data Science and Artificial Intelligence: expanding the use case of High Performance Computing

Abstract:
Over the last 30 years, technical computing has allowed major improvements for product design which became faster, safer and at a better cost. The Internet of Things makes massive amount of data available to be further exploitated. Maintaining the pace of innovation through IT transformation and the capacity to integrate simulation, data management and in some cases AI and ML capacities, are key for industries in order to access new markets and stay competitive. The sessions will focus on different technologies underpinning this IT tranformation, and how to add value for engineering, operation and service.

Noor Shaker (GTN, UK)

Before co-founding GTN, Noor was an assistant professor at Aalborg University in Copenhagen working on machine learning with special interest in generative models. She has more than 50 publications and 1200+ citations. She has co-authored a book on generative methods and has won a number of grants and awards for her research. She served as the Chair of the IEEE Games Technical committee and co-organised task forces and conferences.

Title:
Generative Tensorial Networks: A Quantum Leap in Drug Discovery

Abstract:
Bringing a single new drug to the market costs $2.6bn, often delivering an intervention that barely differs from those already in the market. Projections into the future are not promising, with an expected 50% drop in R&D output every nine years. At GTN we are overcoming the main bottlenecks in drug discovery with our unique patented technology, Generative Tensorial Networks. Our technology combines and builds upon techniques from machine learning and quantum physics, halving development costs and identifying substantially innovative drugs.

Sofia Olheda (UCL, UK)

Sofia is a professor of Statistics, an honorary professor of Computer Science and a senior research associate of Mathematics at University College London. She joined UCL in 2007, before which she was a senior lecturer of statistics (associate professor) at Imperial College London (2006-2007), a lecturer of statistics (assistant professor) (2002-2006), where she also completed her PhD in 2003 and MSci in 2000. She has held three research fellowships while at UCL: UK Engineering and Physical Sciences Springboard fellowship as well as a five-year Leadership fellowship, and now holds a European Research Council Consolidator fellowship. Sofia has contributed to the study of stochastic processes; time series, random fields and networks. She is on the ICMS Programme Committee since September 2008, a member of the London Mathematical Society Research Meetings Committee, a member of the London Mathematical Society Research Policy Committee and an associate Editor for Transactions in Mathematics and its Applications. Sofia was also a member of the Royal Society and British Academy Data Governance Working Group, and the Royal Society working group on machine learning.”

Ulrike Tillmann (Turing Fellow and University of Oxford, UK)

Prof. Ulrike Tillmann FRS has been at the University of Oxford since 1992. She is an algebraic topologist, known in particular for her work on Riemann surfaces and the homology of their moduli spaces. She has long standing research interest in homology stability questions. In 2011 she introduced an annual course (with Abramsky) in Computational Algebraic Topology at masters level. In the last year, Tillmann has co-organized four workshops on topological data analysis, as well as an CMI-LMS research school.

She held an EPSRC Advanced Fellowship 1997-2003. She was invited to present at the ICM in 2002 and was a member of the topology subject panel for both the 2010 and 2014 ICMs. In 2008 she was made a Fellow of the Royal Society and received the Bessel Forschungspreis from the Humboldt Gesellschaft. She is an inaugural Fellow of the AMS.

Title:
The shape of data

Abstract:
The abundance of data poses new challenges. In this talk I will explain how topology, a branch of abstract mathematics, can help to extract new information and get a deeper understanding.