Home
Videos uploaded by user “The Alan Turing Institute”
What is data science?
 
02:50
Data science has emerged in response to the growing need to make sense of the billions of gigabytes of data generated globally every day, and their powerful impact on our society, economy and way of life. But what exactly is it? And why should we care? In this film, we talk to the social scientists, mathematicians, statisticians, computer scientists, engineers, and industry and government thinkers who are driving this new area of research about what data science means to them – and why it is important to all of us. With thanks to: Douglas Flint, HSBC James Geddes, The Alan Turing Institute Ruth King, University of Edinburgh Vidhi Lalchand, Turing/University of Cambridge Ioanna Manolopoulos, UCL Suzy Moat, University of Warwick Jim Smith, University of Warwick Charles Sutton, University of Edinburgh Mark Walport, HM Government Frank Wood, University of Oxford Created with itdrewitself.com/ Find out more at turing.ac.uk.
The Problem of Governance in Distributed Ledger Technologies - Professor Vili Lehdonvirta, OII
 
25:23
Prof. Vili Lehdonvirta is an Associate Professor, Senior Research Fellow, and DPhil Programme Director at the Oxford Internet Institute, University of Oxford, and a Faculty Fellow of the Alan Turing Institute. Previously he worked at London School of Economics, University of Tokyo, and Helsinki Institute for Information Technology. Before his academic career, he worked as a game programmer. He has advised companies, startups, and policy makers in the United States, Europe, and Japan, including Rovio, Warner Brothers, and the World Bank. RESEARCH His research deals with the design and socioeconomic implications of digital marketplaces and platforms, using conventional social research methods and novel data science approaches. He has a PhD in Economic Sociology from Turku School of Economics (2009) and an MSc in Information Networks from Helsinki University of Technology (2005).
The Role of Multi-Agent Learning in Artificial Intelligence Research at DeepMind
 
01:02:12
Event Blurb: In computer science, an agent can be thought of as a computational entity that repeatedly perceives the environment, and takes action so as to optimize long term reward. We consider intelligence to be the ability of an agent to achieve goals in a wide range of environments (Legg & Hutter). Thinking in evolutionary/ecological terms, the richest environments for a given agent are themselves evolving collections of agents. These could be biological organisms, or companies within a given market. In this lecture, Thore will discuss the important role multi-agent learning has to play in artificial intelligence research and the challenges it presents. Specifically, he will discuss two example projects from multi-agent learning work at DeepMind. Firstly, Thore will show how to use advances in deep reinforcement learning to study the age-old question of how cooperation arises among self-interested agents. By defining Sequential Social Dilemmas, this work goes beyond simple matrix games such as the famous game theory example of the Prisoner’s Dilemma, and can model new aspects of social dilemmas such as temporal dynamics and coordination problems. Secondly, Thore will discuss the AlphaGo project, in which DeepMind used the multi-agent algorithm of Learning from Self-Play to create the first computer program to beat a top professional Go player at the full-size game of Go, a feat thought to be at least a decade away by Go and AI experts alike. #aiattheturing #TuringLectures
Professor Stéphane Mallat: "High-Dimensional Learning and Deep Neural Networks"
 
59:34
The Turing Lectures: Mathematics - Professor Stéphane Mallat: High-Dimensional Learning and Deep Neural Networks Click the below timestamps to navigate the video. 00:00:07 Welcome by Professor Andrew Blake, Director, The Alan Turing Institute 00:01:36 Introduction by Professor Jared Tanner 00:03:21 Professor Stéphane Mallat: High-Dimensional Learning and Deep Neural Networks 00:49:53 Q&A The first set of Turing Lectures took place on March 2 with a focus on Mathematics one of the foundations of Data Science. An exciting pair of lectures were delivered by Professors Stéphane Mallat and Mark Newman who considered recent advances in Data Science from a mathematical perspective. Deep Learning and Complex Networks have made the headlines in the scientific and popular press of late, and this Turing Lecture event provided an overview of some of the most recent influential advances in both of these areas. The Alan Turing Institute is the UK's National Institute for Data Science. The Institute’s mission is to: undertake data science research at the intersection of computer science, mathematics, statistics and systems engineering; provide technically informed advice to policy makers on the wider implications of algorithms; enable researchers from industry and academia to work together to undertake research with practical applications; and act as a magnet for leaders in academia and industry from around the world to engage with the UK in data science and its applications. The Institute is headquartered at The British Library, at the heart of London’s knowledge quarter, and brings together leaders in advanced mathematics and computing science from the five founding universities and other partners. Its work is expected to encompass a wide range of scientific disciplines and be relevant to a large number of business sectors. For more information, please visit: https://turing.ac.uk #TuringLectures
Probabilistic Models for Integration Error in Assessment of Functional Cardiac Models
 
03:30
In this video, Dr Chris Oates from the Lloyds-Turing programme on data centric engineering talks about his collaborative research, aiming to build better models of the human heart. These models are being used to study the mechanisms that underlie heart disease and could soon be used in the clinic to help select patient treatment. #datacentricengineering
Towards ambient intelligence in AI-assisted healthcare spaces - Dr Fei-Fei Li, Stanford University
 
41:59
Abstract: Artificial intelligence has begun to impact healthcare in areas including electronic health records, medical images, and genomics. But one aspect of healthcare that has been largely left behind thus far is the physical environments in which healthcare delivery takes place: hospitals, clinics, and assisted living facilities, among others. In this talk I will discuss our work on endowing healthcare spaces with ambient intelligence, using computer vision-based human activity understanding in the healthcare environment to assist clinicians with complex care. I will first present pilot implementations of AI-assisted healthcare spaces where we have equipped the environment with visual sensors. I will then discuss our work on human activity understanding, a core problem in computer vision. I will present deep learning methods for dense and detailed recognition of activities, and efficient action detection, important requirements for ambient intelligence, and I will discuss these in the context of several clinical applications. Finally, I will present work and future directions for integrating this new source of healthcare data into the broader clinical data ecosystem. Bio: Fei-Fei Li is a Professor in the Computer Science Department at Stanford, and the Director of the Stanford Artificial Intelligence Lab. In 2017, she also joined Google Cloud as Chief Scientist of AI and Machine Learning. Dr. Li’s main research areas are in machine learning, deep learning, computer vision and cognitive and computational neuroscience. She has published almost 200 scientific articles in top-tier journals and conferences, including Nature, PNAS, Journal of Neuroscience, New England Journal of Medicine, CVPR, ICCV, NIPS, ECCV, IJCV, and IEEE-PAMI. Dr. Li obtained her B.A. degree in physics from Princeton in 1999 with High Honors, and her PhD degree in electrical engineering from California Institute of Technology (Caltech) in 2005. She joined Stanford in 2009 as an assistant professor, and was promoted to associate professor with tenure in 2012. Prior to that, she was on faculty at Princeton University (2007-2009) and University of Illinois Urbana-Champaign (2005-2006). Dr. Li is the inventor of ImageNet and the ImageNet Challenge, a critical large-scale dataset and benchmarking effort that has contributed to the latest developments in computer vision and deep learning in AI. In addition to her technical contributions, she is a national leading voice for advocating diversity in STEM and AI. She is a co-founder of Stanford’s renowned SAILORS outreach program for high school girls and the national non-profit AI4ALL. For her work in AI, Dr. Li has been a keynote speaker at the Grace Hopper Conference in 2017 and TED2015 main conference. She is a recipient of the 2017 Athena Academic Leadership Award, IAPR 2016 J.K. Aggarwal Prize, the 2016 NVIDIA Pioneer in AI Award, 2014 IBM Faculty Fellow Award, 2011 Alfred Sloan Faculty Award, 2012 Yahoo Labs FREP award, 2009 NSF CAREER award, the 2006 Microsoft Research New Faculty Fellowship, and a number of Google Research awards. Work from Dr. Li’s lab have been featured in a variety of popular press magazines and newspapers including New York Times, Wall Street Journal, Fortune Magazine, Science, Wired Magazine, MIT Technology Review, Financial Times, and more. She was selected as a 2017 Women in Tech by the ELLE Magazine, a 2017 Awesome Women Award by Good Housekeeping, a Global Thinker of 2015 by Foreign Policy, and one of the “Great Immigrants: The Pride of America” in 2016 by the Carnegie Foundation (past winners include Albert Einstein, Yo-Yo Ma, Sergey Brin and more). #TuringSeminars
Learning Explanatory Rules from Noisy Data - Richard Evans, DeepMind
 
26:09
Artificial Neural Networks are powerful function approximators capable of modelling solutions to a wide variety of problems, both supervised and unsupervised. As their size and expressivity increases, so too does the variance of the model, yielding a nearly ubiquitous overfitting problem. Although mitigated by a variety of model regularisation methods, the common cure is to seek large amounts of training data—which is not necessarily easily obtained—that sufficiently approximates the data distribution of the domain we wish to test on. In contrast, logic programming methods such as Inductive Logic Programming offer an extremely data-efficient process by which models can be trained to reason on symbolic domains. However, these methods are unable to deal with the variety of domains neural networks can be applied to: they are not robust to noise in or mislabelling of inputs, and perhaps more importantly, cannot be applied to non-symbolic domains where the data is ambiguous, such as operating on raw pixels. In this paper, we propose a Differentiable Inductive Logic framework (∂ILP), which can not only solve tasks which traditional ILP systems are suited for, but shows a robustness to noise and error in the training data which ILP cannot cope with. Furthermore, as it is trained by back-propagation against a likelihood objective, it can be hybridised by connecting it with neural networks over ambiguous data in order to be applied to domains which ILP cannot address, while providing data efficiency and generalisation beyond what neural networks on their own can achieve. The workshop was held on January 11th and 12th. Logic has proved in the last decades a powerful tool in understanding complex systems. It is instrumental in the development of formal methods, which are mathematically based techniques obsessing on hard guarantees. Learning is a pervasive paradigm which has seen tremendous success recently. The use of statistical approaches yields practical solutions to problems which yesterday seemed out of reach. These two mindsets should not be kept apart, and many efforts have been made recently to combine the formal reasoning offered by logic and the power of learning. The goal of this workshop is to bring together expertise from various areas to try and understand the opportunities offered by combining logic and learning. There are 12 invited speakers and a light programme (less than 5h per day) so as to give enough time for discussions.
Professor Mihaela van der Schaar, Oxford University
 
28:18
Mihaela van der Schaar is MAN Professor of Quantitative Finance in the Oxford – MAN Institute (OMI) and the Department of Engineering Science at Oxford, and Fellow of Christ Church College The current emphasis of her research is on machine learning with applications to medicine, finance and education. She has also worked on data science, network science, game theory, signal processing, communications, and multimedia. Prior to her academic career, she was a Senior Researcher at Philips Research in the Netherlands and USA; from this work she holds 33 patents. #TuringShortTalks
Professor Luciano Floridi: "Ethics in the Age of Information"
 
01:14:47
The Turing Lectures: Social Science and Ethics - Professor Luciano Floridi, Oxford Internet Institute, University of Oxford: “Ethics in the Age of Information” Click the below timestamps to navigate the video. 00:00:07 Introduction by Professor Andrew Blake, Director, The Alan Turing Institute 00:02:20 Professor Luciano Floridi, Oxford Internet Institute, University of Oxford: “Ethics in the Age of Information” 00:59:05 Q&A The excitement of Data Science brings the need to consider the ethics associated with the information age. Likewise a revolution in political science is taking place where the internet, social media and real time electronic monitoring has brought about increased mobilisation of political movements. In addition the generation of huge amounts of data from such processes presents on the one hand opportunities to analyse and indeed predict political volatility, and on the other ethical and technical challenges which will be explored by two of the foremost philosophers and political scientists. The Alan Turing Institute is the UK's National Institute for Data Science. The Institute’s mission is to: undertake data science research at the intersection of computer science, mathematics, statistics and systems engineering; provide technically informed advice to policy makers on the wider implications of algorithms; enable researchers from industry and academia to work together to undertake research with practical applications; and act as a magnet for leaders in academia and industry from around the world to engage with the UK in data science and its applications. The Institute is headquartered at The British Library, at the heart of London’s knowledge quarter, and brings together leaders in advanced mathematics and computing science from the five founding universities and other partners. Its work is expected to encompass a wide range of scientific disciplines and be relevant to a large number of business sectors. For more information, please visit: https://turing.ac.uk #TuringLectures
Cedric Villani - Franco-British Artificial Intelligence Conference
 
17:22
France and the UK have both taken significant steps to take full advantage of the AI revolution, and a new conference on 16 January aims to catalyse cooperation between these two leading AI nations. Organised by The Alan Turing Institute and London French Tech Hub, the first Franco-British conference on AI will bring together French and UK leading incubators, investors, entrepreneurs and policymakers. Cédric Villani (French MP and 2010 Fields Medal winner) and Jérôme Pesenti (CEO, Benevolent Tech), major actors in both countries’ AI strategies, will introduce this unique conference, in which the AI start-up communities will have the opportunity to interact with government, academics and industry and for each to learn from the other. Both the French and UK governments have shown a commitment to harnessing the positive effects of AI as one of the fastest-growing tech sectors. The UK Industrial Strategy, presented in November by Business Secretary Greg Clark, puts great emphasis on AI, and one month after Dame Wendy Hall and Jérôme Pesenti published the independent report “Growing the artificial intelligence industry in the UK”. Critically, the UK government took up this report’s recommendation that The Alan Turing Institute becomes the national research centre for AI, adding to the data science mission it was created with. Across the Channel, the French Prime Minister asked Cédric Villani in September 2017 to head a special AI taskforce to establish a French national strategy on AI. This taskforce’s conclusions will be revealed in early 2018. Other leaders in AI will discuss the two start-up ecosystems, funding environments and academic sectors. The first panel will present both research landscapes and ways to improve academic cooperation. The second session will present the two AI markets and investment trends, exploring ways to foster new partnerships between the UK and France. This is a vital moment to bring together leading scientists, entrepreneurs and institutional actors from France and the UK to compare best practice and identify ways to work together to produce an environment in which AI can flourish. #aiattheturing
Turing Lecture: Data Science for Medicine
 
01:20:59
Medicine 2.0: Using Machine Learning to Transform Medical Practice and Discovery In this talk, Mihaela will present her view of the transformation of medicine through the use of machine learning, and some of her own contributions. This transformation is already being felt in every aspect of medicine: from clinical support for personalized diagnosis and prognosis to the estimation of individualized treatment effects without the need for clinical trials to medical discovery to the entire path of patient care. The heart of this transformation is the intelligent use of existing data. Because of the unique and complex characteristics of medical data and medical questions, many familiar methods are inadequate. Mihaela’s work develops novel machine learning methods and applies them to a wide variety of medical settings. This work achieves enormous improvements over current clinical practice and over existing machine learning methods. This talk will explore some of Mihaela’s work and vision of how data science can transform medical discovery and care. Biography Mihaela van der Schaar is the Man Professor, Department of Engineering Science, University of Oxford and Faculty Fellow at The Alan Turing Institute. Her main research interest is on machine learning and artificial intelligence applied to medicine. She is an Institute of Electrical and Electronics Engineers (IEEE) Fellow (2009) and has been a Distinguished Lecturer of the Communications Society, the Editor in Chief of IEEE Transactions on Multimedia, and member of the Senior Editorial Board member of IEEE Journal on Emerging and Selected Topics in Circuits and Systems (JETCAS) and IEEE Journal on Selected Topics in Signal Processing (JSTSP). She received an NSF CAREER Award (2004), the Best Paper Award from IEEE Transactions on Circuits and Systems for Video Technology (2005), the Okawa Foundation Award (2006), the IBM Faculty Award (2005, 2007, 2008), the Most Cited Paper Award from EURASIP: Image Communications Journal (2006), the Gamenets Conference Best Paper Award (2011) and the 2011 IEEE Circuits and Systems Society Darlington Best. #TuringLectures
Professor Mark Newman: "Epidemics, Erdos numbers, and the Internet"
 
55:28
The Turing Lectures: Mathematics - Professor Mark Newman: "Epidemics, Erdos numbers, and the Internet" Click the below timestamps to navigate the video. 00:00:07 Lecture introduction by Professor Jared Tanner 00:01:14 Professor Mark Newman: Epidemics, Erdos numbers, and the Internet: The Form and Function of Networks 00:51:02 Q&A The first set of Turing Lectures took place on March 2 2016 with a focus on Mathematics one of the foundations of Data Science. An exciting pair of lectures were delivered by Professors Stéphane Mallat and Mark Newman who considered recent advances in Data Science from a mathematical perspective. Deep Learning and Complex Networks have made the headlines in the scientific and popular press of late, and this Turing Lecture event provided an overview of some of the most recent influential advances in both of these areas. For more information, please visit: https://turing.ac.uk/turing-lectures-... The Alan Turing Institute is the UK's National Institute for Data Science. The Institute’s mission is to: undertake data science research at the intersection of computer science, mathematics, statistics and systems engineering; provide technically informed advice to policy makers on the wider implications of algorithms; enable researchers from industry and academia to work together to undertake research with practical applications; and act as a magnet for leaders in academia and industry from around the world to engage with the UK in data science and its applications. The Institute is headquartered at The British Library, at the heart of London’s knowledge quarter, and brings together leaders in advanced mathematics and computing science from the five founding universities and other partners. Its work is expected to encompass a wide range of scientific disciplines and be relevant to a large number of business sectors. For more information, please visit: https://turing.ac.uk #TuringLectures
Simulation: The Challenge for Data Science
 
01:01:04
While machine learning has recently had dramatic successes, there is a large class of problems that it will never be able to address on its own. To test a policy proposal, for example, often requires understanding a counterfactual scenario that has never existed in the past, and that may fundamentally alter the statistical properties of data in the future. Simulation models provide an alternative, in which one incorporates much stronger prior knowledge about structure and causal interaction. Making such models quantitatively accurate enough that they can be trusted for policy analysis poses difficult challenges for parameter estimation and initialization. Simulation models are usually formulated in terms of micro-states, such as individual households, whereas measurements are often only available at an aggregate level, such as GDP or unemployment. To run a simulation as a time series model to forecast macro-states requires initializing the micro-states of the model to match macroscopic data, and can create severe problems if not done carefully. Borrowing from methods of data assimilation used in meteorology, we introduce a new method for solving this problem. I will review the literature on parameter estimation for simulation models and discuss the relevant challenges and the opportunities to create a fusion with machine learning. My main point is that the problems of parameter estimation and initialization for simulation models are at the cutting edge of data science. These problems are ripe to be solved, and their solution will catapult simulation science to a level of usefulness similar to that of machine learning. Work in this area should figure prominently on the agenda of The Alan Turing Institute. Doyne Farmer is Director of the Complexity Economics program at the Institute for New Economic Thinking at the Oxford Martin School, Professor in the Mathematical Institute at the University of Oxford, and an External Professor at the Santa Fe Institute. His current research is in economics and finance, including agent-based modeling, financial instability and technological progress. He was a founder of Prediction Company, a quantitative automated trading firm that was sold to the United Bank of Switzerland in 2006. His past research includes complex systems, dynamical systems theory, time series analysis and theoretical biology. During the eighties he was an Oppenheimer Fellow and the founder of the Complex Systems Group at Los Alamos National Laboratory. While a graduate student in the 70’s he built the first wearable digital computer, which was successfully used to predict the game of roulette. #TuringSeminars
A review of machine learning techniques for anomaly detection - Dr David Green
 
21:46
The Alan Turing Institute, headquartered in the British Library, London, was created as the national institute for data science in 2015. In 2017, as a result of a government recommendation, we added artificial intelligence to our remit. The Institute is named in honour of Alan Turing (23 June 1912 – 7 June 1954), whose pioneering work in theoretical and applied mathematics, engineering and computing are considered to be the key disciplines comprising the fields of data science and artificial intelligence. Five founding universities – Cambridge, Edinburgh, Oxford, UCL and Warwick – and the UK Engineering and Physical Sciences Research Council created The Alan Turing Institute in 2015. Eight new universities – Leeds, Manchester, Newcastle, Queen Mary University of London, Birmingham, Exeter, Bristol, and Southampton – are set to join the Institute in 2018.# #TuringSeminars
Poisson random fields for dynamic feature models: Valerio Perrone, Oxford-Warwick Stats Programme
 
29:54
This talk is based on the article: http://jmlr.org/papers/volume18/16-541/16-541.pdf In a feature allocation model, each data point depends on a collection of unobserved latent features. For example, we might classify a corpus of texts by describing each document via a set of topics; the topics then determine a distribution over words for that document. In a Bayesian nonparametric setting, the Indian Buffet Process (IBP) is a popular prior in which the number of topics is unknown a priori. However, the IBP is static in that it does not account for the change in popularity of topics over time. This talk introduces the Wright-Fisher Indian Buffet Process (WF-IBP), a probabilistic model for collections of time-stamped documents. This is applied to develop a nonparametric focused topic model for collections of time-stamped text documents, and explore the full corpus of NIPS papers published from 1987 to 2015. Bio: Valerio Perrone is a final year PhD student at the Oxford-Warwick Statistics Programme (OxWaSP), working under the joint supervision of Professor Yee Whye Teh (Oxford), Dr. Dario Spanò (Warwick), and Dr. Paul Jenkins (Warwick). His research interests lie in the fields of Bayesian machine learning and deep learning. In particular, he is interested in developing algorithms for large-scale machine learning and Bayesian nonparametric models with realistic dependency structures. Applications of his work include topic modelling, recommender systems and population genetics. #TuringSeminars
Celebrating getting started
 
02:29
The Alan Turing Institute was founded in 2015 by the universities of Cambridge, Edinburgh, Oxford, Warwick, and UCL and the EPSRC with the mission to advance the world-changing potential of data science. This autumn we marked the start of our first ever academic year with an event bringing together staff, researchers and partners to raise a glass to the year ahead. Here's a short video celebrating being 'open for business'. For news, events and how to get involved, visit turing.ac.uk. With thanks to ItDrewItself http://itdrewitself.com/.
Artificial Intelligence: GDPR and beyond - Dr. Sandra Wachter, University of Oxford
 
18:26
Dr. Sandra Wachter is a lawyer and Research Fellow in Data Ethics, AI, robotics and Internet Regulation/cyber-security at the Oxford Internet Institute and the Alan Turing Institute in London as well as a member of the Law Committee of the IEEE. She serves as a policy advisor for governments and NGO’s around the world on regulatory and ethical questions concerning emerging technologies. Prior to joining the OII, Sandra worked at the Royal Academy of Engineering and at the Austrian Ministry of Health. Event information: The General Data Protection Regulation (GDPR) will come into force in the European Union in May 2018. The regulation is designed to strengthen the European data protection regime for personal data of EU residents which is processed within the EU and outside it. Notably, the GDPR also addresses the use of automated algorithmic decision-making and profiling in processing personal data, raising questions about the extent to which data subjects have rights to meaningful information about the logic involved, to obtain human intervention, and to contest decisions made by solely automated systems. By addressing a wide array of concepts, including fairness, transparency, privacy, consent, and interpretability, the GDPR is set to reshape the relationship between governments, corporations, and the individuals whose personal data they process. Since organisations found not to be in compliance with the regulation will face serious penalties (up to 4% of global revenue), there is great interest in exactly what the GDPR does and does not require, as well as how it will be interpreted after implementation. This day-long, expert-led workshop will explain the purposes and provisions of the GDPR, and explore what next steps might be for the regulation of artificial intelligence.
End-to-End Differentiable Proving: Tim Rocktäschel, University of Oxford
 
42:13
We introduce neural networks for end-to-end differentiable proving of queries to knowledge bases by operating on dense vector representations of symbols. These neural networks are constructed recursively by taking inspiration from the backward chaining algorithm as used in Prolog. Specifically, we replace symbolic unification with a differentiable computation on vector representations of symbols using a radial basis function kernel, thereby combining symbolic reasoning with learning subsymbolic vector representations. By using gradient descent, the resulting neural network can be trained to infer facts from a given incomplete knowledge base. It learns to (i) place representations of similar symbols in close proximity in a vector space, (ii) make use of such similarities to prove queries, (iii) induce logical rules, and (iv) use provided and induced logical rules for multi-hop reasoning. We demonstrate that this architecture outperforms ComplEx, a state-of-the-art neural link prediction model, on three out of four benchmark knowledge bases while at the same time inducing interpretable function-free first-order logic rules.
Professor Mark Girolami: "Probabilistic Numerical Computation: A New Concept?"
 
01:01:07
The Turing Lectures: The Intersection of Mathematics, Statistics and Computation - Professor Mark Girolami: "Probabilistic Numerical Computation: A New Concept?" Click the below timestamps to navigate the video. 00:00:09 Introduction by Professor Jared Tanner 00:01:38 Professor Mark Girolami: "Probabilistic Numerical Computation: A New Concept?" 00:54:48 Q&A Lecture blurb: The vast amounts of data in many different forms becoming available to politicians, policy makers, technologists, and scientists of every hue presents tantalising opportunities for making advances never before considered feasible. Yet with these apparent opportunities has come an increase in the complexity of the mathematics required to exploit this data. These sophisticated mathematical representations are much more challenging to analyse, and more and more computationally expensive to evaluate. This is a particularly acute problem for many tasks of interest, such as making predictions since these will require the extensive use of numerical solvers for linear algebra, optimization, integration or differential equations. These methods will tend to be slow, due to the complexity of the models, and this will potentially lead to solutions with high levels of uncertainty. This talk will introduce our contributions to an emerging area of research defining a nexus of applied mathematics, statistical science and computer science, called “probabilistic numerics”. The aim is to consider numerical problems from a statistical viewpoint, and as such provide numerical methods for which numerical error can be quantified and controlled in a probabilistic manner. This philosophy will be illustrated on problems ranging from predictive policing via crime modelling to computer vision, where probabilistic numerical methods provide a rich and essential quantification of the uncertainty associated with such models and their computation. Bio After graduation from the University of Glasgow, Mark Girolami spent the first ten years of his career with IBM as an Engineer. After this he undertook, on a part time basis, a PhD in Statistical Signal Processing whilst working in a Scottish Technical College. He then went on rapidly to hold senior professorial positions at the University of Glasgow, and University College London. He is an EPSRC Established Career Research Fellow (2012 - 2017) and previously an EPSRC Advanced Research Fellow (2007 - 2012). He is the Director of the EPSRC funded Research Network on Computational Statistics and Machine Learning and in 2011, was elected to the Fellowship of the Royal Society of Edinburgh, when he was also awarded a Royal Society Wolfson Research Merit Award. He has been nominated by the Institute of Mathematical Statistics to deliver a Medallion Lecture at the Joint Statistical Meeting in 2017. He is currently one of the founding Executive Directors of the Alan Turing Institute for Data Science His research and that of his group covers the development of advanced novel statistical methodology driven by applications in the life, clinical, physical, chemical, engineering and ecological sciences. He also works closely with industry where he has several patents leading from his work on e.g. activity profiling in telecommunications networks and developing statistical techniques for the machine based identification of counterfeit currency which is now an established technology used in current Automated Teller Machines. At present he works as a consultant for the Global Forecasting Team at Amazon in Seattle. The Alan Turing Institute is the UK's National Institute for Data Science. The Institute’s mission is to: undertake data science research at the intersection of computer science, mathematics, statistics and systems engineering; provide technically informed advice to policy makers on the wider implications of algorithms; enable researchers from industry and academia to work together to undertake research with practical applications; and act as a magnet for leaders in academia and industry from around the world to engage with the UK in data science and its applications. The Institute is headquartered at The British Library, at the heart of London’s knowledge quarter, and brings together leaders in advanced mathematics and computing science from the five founding universities and other partners. Its work is expected to encompass a wide range of scientific disciplines and be relevant to a large number of business sectors. For more information, please visit: https://turing.ac.uk #TuringLectures
Meltdown and Spectre - Professor Mark Handley, UCL
 
01:43:02
The Meltdown and Spectre vulnerabilities in almost all modern CPUs have received a great deal of publicity in the last week. Operating systems and hypervisors need significant changes to how memory management is performed, CPU firmware needs updating, compilers are being modified to avoid risky instruction sequences, and browsers are being patched to prevent scripts having access to accurate time. All this because of how speculative execution is handled in modern pipelined superscalar CPUs, and how side-channel attacks reveal information about execution that the CPU tries to pretend did not happen. Mark Handley will explain what modern CPUs actually do to go fast, discuss how this leads to the Meltdown and Spectre vulnerabilities, and summarize the mitigations that are being put in place. Bio Mark Handley joined the Computer Science department at UCL as Professor of Networked Systems in 2003, receiving a Royal Society-Wolfson Research Merit Award. From 2003-2010 he led the Networks Research Group, which has a long history dating back to 1973 when UCL became the first site outside the United States to join the ARPAnet, which was the precursor to today's Internet. Prior to joining UCL, Professor Handley was based at the International Computer Science Institute in Berkeley, California, where he co-founded the AT&T Center for Internet Research at ICSI (ACIRI). Professor Handley has been very active in the area of Internet Standards, and has served on the Internet Architecture Board, which oversees much of the Internet standardisation process. He is the author of 33 Internet standards documents (RFCs), including the Session Initiation Protocol (SIP), which is the principal way telephony signalling is performed in Internet-based telephone networks. Recently he has been standardizing multipath extensions to TCP. Professor Handley's research interests include the Internet architecture (how the components fit together to produce a coherent whole), congestion control (how to match the load offered to a network to the changing available capacity of the network), Internet routing (how to satisfy competing network providers' requirements, while ensuring that traffic takes a good path through the network), and defending networks against denial-of-service attacks. He also founded the XORP project to build a complete open-source Internet routing software stack. #datascienceclasses
Time Series Modelling and State Space Models: Professor Chris Williams, University of Edinburgh
 
01:35:29
- AR, MA and ARMA models - Parameter estimation for ARMA models - Hidden Markov Models (definitions, inference, learning) - Linear-Gaussian HMMs (Kalman filtering) - More advanced topics (more elaborate state-space models, and recurrent neural networks) #datascienceclasses
Counterfactual Fairness: Matt Kusner, The Alan Turing Institute
 
46:22
Dr Kusner is a Research Fellow at The Alan Turing Institute. He was previously a visiting researcher at Cornell University, under the supervision of Kilian Q Weinberger, and received his PhD in Machine Learning from Washington University in St Louis. His research is in the areas of counterfactual fairness, privacy, budgeted learning, model compression and Bayesian optimisation. Talk title: Counterfactual Fairness Synopsis: Machine learning can impact people with legal or ethical consequences when it is used to automate decisions in areas such as insurance, lending, hiring, and predictive policing. In many of these scenarios, previous decisions have been made that are unfairly biased against certain subpopulations, for example those of a particular race, gender, or sexual orientation. Since this past data may be biased, machine learning predictors must account for this to avoid perpetuating or creating discriminatory practices. Matt will present a framework for modelling fairness using tools from causal inference. This definition of counterfactual fairness captures the intuition that a decision is fair towards an individual if it the same in (a) the actual world and (b) a counterfactual world where the individual belonged to a different demographic group. The framework is demonstrated on a real-world problem of fair prediction of success in law school. #aiattheturing
Artificial Intelligence, ethics and the law: What challenges? What opportunities?
 
55:40
Artificial Intelligence (AI) is no longer sci-fi. From driverless cars to the use of machine learning algorithms to improve healthcare services and the financial industry, AI and algorithms are shaping our daily practices, and a fast-growing number of fundamental aspects of our societies. This can lead to dangerous situations in which vital decision making is automated – for instance in credit scoring, or sentencing – but limited policies exist for citizens subject to such AI technologies embedded in our social institutions to seek redress. Similarly, well-intended technologists might release AI into society that is ethically unsound. A growing body of literature on improving the auditability and transparency of algorithms is being developed. Yet, more is needed to develop a shared understanding about the fundamental issues at the heart of the debate on AI, algorithms, the law, and ethics. These issues taken together are leading to a renewed focus on, and increasing concern about, the ethical and legal impact of AI on our societies. In this panel we bring together five thought leaders on AI from the corporate sector, academia, politics, and civil society to discuss. We will hear from Paul Nemitz, Monica Beltrametti, Alan Winfield, Vidushi Marda and Sandra Wachter, the conversation will be moderated by Corinne Cath. Panel: - Paul Nemitz – Director responsible for Fundamental rights and Union citizenship in the Directorate-General Justice of the European Commission - Monica Beltrametti – Director at NAVER Labs Europe - Alan Winfield – Professor of Robot Ethics University of the West of England - Sandra Wachter – Post-doctoral Researcher at Oxford Internet Institute and The Alan Turing Institute - Vidushi Marda – Policy Advisor with NGO Article 19 #aiattheturing
The GDPR and Beyond: Elizabeth Denham, UK Information Commissioner
 
21:09
Elizabeth Denham was appointed UK Information Commissioner in July 2016, having previously held the position of Information and Privacy Commissioner for British Columbia, Canada and Assistant Privacy Commissioner of Canada.The Commissioner has also demonstrated a focus on the essential role data protection can play in innovation, and the importance of organisations understanding the growing impetus on companies to be accountable for what they do with personal data. This forms a central part of the new General Data Protection Regulation, which comes into force in May 2018. In 2011, Ms Denham was honoured as a UBC distinguished alumni for her pioneering work in archives and leadership in the field of access and privacy. In 2013, she received the Queen Elizabeth II Diamond Jubilee Medal for her service as an Officer of the Legislature of British Columbia, Canada. In 2017, she was recognised as being one of the three most influential people in data-driven business in the annual DataIQ 100 list. She was honoured to accept the appointment of Visiting Professor in University College London’s department of Information Studies. The professorship will extend until 2022. In 2018, she was named as the most influential person in data-driven business in the updated DataIQ 100 list. Event information: The General Data Protection Regulation (GDPR) will come into force in the European Union in May 2018. The regulation is designed to strengthen the European data protection regime for personal data of EU residents which is processed within the EU and outside it. Notably, the GDPR also addresses the use of automated algorithmic decision-making and profiling in processing personal data, raising questions about the extent to which data subjects have rights to meaningful information about the logic involved, to obtain human intervention, and to contest decisions made by solely automated systems. By addressing a wide array of concepts, including fairness, transparency, privacy, consent, and interpretability, the GDPR is set to reshape the relationship between governments, corporations, and the individuals whose personal data they process. Since organisations found not to be in compliance with the regulation will face serious penalties (up to 4% of global revenue), there is great interest in exactly what the GDPR does and does not require, as well as how it will be interpreted after implementation. This day-long, expert-led workshop will explain the purposes and provisions of the GDPR, and explore what next steps might be for the regulation of artificial intelligence.
There is no AI ethics: The human origins of machine prejudice - Dr Joanna Bryson
 
33:49
Bridging disciplines in analysing text as social and cultural data workshop (21-22 September, 2017) The potential benefits of using large-scale text data to study social and cultural phenomena is increasingly being recognized, but researchers are currently scattered across a range of often distinct research communities. However, many methodological challenges cut across research disciplines and require interdisciplinary synergies. This workshop aims to address the gap between research methodologies in NLP/ML and the humanities and the social sciences. More information here: https://dongpng.github.io/attached/ #aiattheturing
Statistical Methodology and Theory : Professor Richard Samworth, Cambridge
 
02:21:01
Statistical Methodology and Theory with Richard Samworth (University of Cambridge, Turing Faculty Fellow) Richard Samworth is Professor of Statistics in the Statistical Laboratory at the University of Cambridge and a Fellow of St John’s College. He received his PhD, also from the University of Cambridge, in 2004, and currently holds an EPSRC Early Career Fellowship. Research His main research interests are in developing methodology and theory for high-dimensional and nonparametric statistical inference. He is currently particularly interested in techniques for handling statistical challenges in Big Data that rely on perturbations of the data and aggregation. Examples include random projection ensembles for high-dimensional classification and subsampling for variable selection. Other interests include shape-constrained and other nonparametric function estimation problems, Independent Component Analysis, and the bootstrap and its variants (e.g. bagging). #datascienceclasses
Delayed column generation in large scale integer optimization problems - Professor Raphael Hauser
 
02:41:14
Mixed linear integer programming problems play an important role in many applications of decision mathematics, including data science. Algorithms typically solve such problems via a sequence of linear programming approximations and a divide-and-conquer approach (branch-and-bound, branch-and-cut). The simplex algorithm is preferred for the solution of the LP subproblems, due to its ability to take previous computations into account in warm-starts of subproblems with additional constraints. For very large scale problems this approach would be limited by memory constraints, but the so-called delayed column generation approach makes it possible to get away with only holding a small part of the problem parameters in memory and generating the required data on the fly. To make this approach viable, the IP problem needs special structure based on partial decoupling of the decision variables, which is often present in big data problems. After a short recap of the simplex method and branch-and-bound for general integer programming problems, we discuss delayed column generation in the context of the classical cutting stock problem, before discussing the branch-and-price method in a general setup.
Professor Mike West: Structured Dynamic Graphical Models & Scaling Multivariate Time Series
 
01:13:46
The Turing Lectures - Professor Mike West: Structured Dynamic Graphical Models & Scaling Multivariate Time Series. Click the below timestamps to navigate the video. 00:00:12 Welcome & Introduction by Doctor Ioanna Manolopoulou 00:01:19 Professor Mike West: Structured Dynamic Graphical Models & Scaling Multivariate Time Series 01:01:47 Q&A Lecture blurb: Title: Structured Dynamic Graphical Models & Scaling Multivariate Time Series Abstract: Mike West, visiting speaker from Duke University, discusses some of their recent R&D with dynamic statistical models for multivariate time series forecasting that represents a shift in modelling approaches in response to the coupled challenges of scalability and model complexity. Building “simple” and computationally tractable models of univariate time series is a starting point. Decouple/Recouple is an overlaid strategy for coherent Bayesian analysis: That is, “decouple” a high-dimensional system into the lowest-level components for simple/fast analysis; and then, “recouple”– on a sound theoretical basis– to rebuild the larger multivariate process for full/formal/coherent inferences and predictions. He will discuss Bayesian dynamic dependency networks (DDNs) and the broader class of simultaneous graphical dynamic linear models (SGDLMs) that define a framework to address these goals. Aspects of model specification, fitting and computation include importance sampling and variational Bayes methods to implement sequential analysis and forecasting. Studies in financial time series forecasting and portfolio decisions highlight the utility of the models. The advances in Bayesian dynamic modelling– and in thinking about coherent and implementable strategies for scalability to higher-dimensions (i.e. to “big, dynamic data”)– are nicely exemplified in these contexts. Aspects of this talk represent recent joint work with: Zoey Zhao, 2013 PhD at Duke University, now at Citadel llc, Chicago; Lutz Gruber, 2015 PhD at the Technical University of Munich, now at Quantco, Cologne; and Meng Amy Xie, 2012 BS at Duke University, and current PhD student in Statistical Science at Duke. Bio: Mike West holds a Duke University distinguished chair as the Arts & Sciences Professor of Statistics & Decision Sciences in the Department of Statistical Science, where he led the development of statistics from 1990-2002. A past president of the International Society for Bayesian Analysis (ISBA), West has served the international statistics profession in founding roles for ISBA, the National Institute of Statistical Sciences and the Statistical & Applied Mathematical Sciences Institute in the USA, and on advisory boards of national research institutes in UK and Japan, among other professional activities. West works in Bayesian statistical methods and application, with over 180 papers, 3 books and several edited volumes related to core statistics and interdisciplinary applications in business, econometrics and finance, sig-nal processing, climatology, public health, genomics, immunology, neurophysiology, systems biology and other areas. West has received a number of international awards for research and professional service, including the international Mitchell Prize for research in applied Bayesian statistics (three times), the American Statistical As-sociation JASA award (twice), the Zellner Medal of ISBA (in 2014), and multiple distinguished speaking awards. He has been a statistical consultant for multiple companies, banks, government agencies and academic centers, co-founder of a successful biotech company, and past or current advisor or board member for several financial and IT companies. West teaches broadly in Bayesian statistics, in academia and through short-courses, works with and advises many undergraduates and Master’s students, and has mentored more than 55 primary PhD students and postdoctoral associates, most of whom are now in academic, industrial or governmental positions involving advanced statistical research. The Alan Turing Institute is the UK's National Institute for Data Science. The Institute is headquartered at The British Library, at the heart of London’s knowledge quarter, and brings together leaders in advanced mathematics and computing science from the five founding universities and other partners. Its work is expected to encompass a wide range of scientific disciplines and be relevant to a large number of business sectors. For more information, please visit: https://turing.ac.uk #TuringLectures
Turing Lecture: Algorithmic Accountability: Professor Ben Shneiderman, University of Maryland
 
01:21:33
Algorithmic Accountability: Designing for safety through human-centered independent oversight In this talk, Ben Shneiderman will explore how some social strategies can play a powerful role in making systems more reliable and trustworthy. He will look at strategies that support human-centred independent oversight during planning, continuous monitoring during operation, and retrospective analyses following failures. Vital services such as communications, financial trading, healthcare, and transportation, depend on sophisticated algorithms. Some of these rely on unpredictable artificial intelligence techniques that are increasingly embedded in complex software systems, such as deep learning. As high-speed trading, medical devices, and autonomous aircraft become more widely implemented, stronger checks become necessary to prevent failures. Ben will discuss how by designing strategies that promote human-centred systems which are comprehensible, predictable and controllable we can increase safety and make failure investigations more effective. He will also stress the importance of clarifying responsibility for failures to stimulate improved design thinking. Biography: Ben Shneiderman is a Distinguished University Professor in the Department of Computer Science, Founding Director (1983-2000) of the Human-Computer Interaction Laboratory, and a Member of the UM Institute for Advanced Computer Studies (UMIACS) at the University of Maryland. He is a Fellow of the AAAS, ACM, IEEE, and NAI, and a Member of the National Academy of Engineering, in recognition of his pioneering contributions to human-computer interaction and information visualisation. His contributions include the direct manipulation concept, clickable highlighted web-links, touchscreen keyboards, dynamic query sliders, development of treemaps, novel network visualisations for NodeXL, and temporal event sequence analysis for electronic health records. Shneiderman is the lead author of Designing the User Interface: Strategies for Effective Human-Computer Interaction (6th ed., 2016). He co-authored Readings in Information Visualization: Using Vision to Think with Stu Card and Jock Mackinlay and Analyzing Social Media Networks with NodeXL with Derek Hansen and Marc Smith. #TuringLectures
Dr Vidit Nanda, University of Oxford
 
19:27
Bio Dr Vidit Nanda is an applied and computational algebraic topologist. Before starting at the Alan Turing Institute, he worked a post-doctoral research fellow at the University of Pennsylvania and as a PhD candidate in mathematics at Rutgers University. http://www.sas.upenn.edu/~vnanda/ Research Dr Nanda develops algebraic-topological theories, algorithms and software for the analysis of non-linear data and complex systems arising in various scientific contexts. In particular, he employs discrete Morse-theoretic techniques to substantially compress cell complexes built around the input data without modifying core topological properties. His recent work has involved excursions into computational geometry, cellular sheaf theory and higher-categorical localization. #TuringShortTalks
“The Automatic Statistician”– Professor Zoubin Ghahramani
 
54:56
Talk given by Professor of Information Engineering at the University of Cambridge, leader of the Cambridge Machine Learning Group, and the Cambridge Liaison Director of the Alan Turing Institute; Zoubin Ghahramani. The lecture regards the use of Bayesian model selection strategies that automatically select and use models to generate human readable reports. #TuringSeminars
Fellow Short Talks: Professor Zoubin Ghahramani, University of Cambridge
 
27:58
Bio Zoubin Ghahramani FRS is Professor of Information Engineering at the University of Cambridge, where he leads the Machine Learning Group, and The Alan Turing Institute’s University Liaison Director for Cambridge. He is also the Deputy Academic Director of the Leverhulme Centre for the Future of Intelligence, and a Fellow of St John’s College Cambridge. He has worked and studied at the University of Pennsylvania, MIT, the University of Toronto, the Gatsby Unit at UCL, and CMU. He is co-founder of Geometric Intelligence and advises a number of AI and machine learning companies. He has served as programme and general chair of the leading international conferences in machine learning: AISTATS, ICML, and NIPS. In 2015 he was elected a Fellow of the Royal Society. Research Zoubin’s current research interests include statistical machine learning, Bayesian nonparametrics, scalable inference, deep learning, probabilistic programming, Bayesian optimisation, and automating data science. His Automatic Statistician project aims to automate the exploratory analysis and modelling of data, discovering good models for data and generating a human-interpretable natural language summary of the analysis. He and his group have also worked on automating inference (though probabilistic programming) and on automating the allocation of computational resources. #aiattheturing #TuringShortTalks
Artificial Intelligence per Kilowatt-hour: Max Welling, University of Amsterdam
 
54:32
Professor Welling is a research chair in Machine Learning at the University of Amsterdam and a Vice President Technologies at Qualcomm. He has a secondary appointment at the Canadian Institute for Advanced Research (CIFAR). He is co-founder of “Scyfer BV” a university spin-off in deep learning which was acquired by Qualcomm. Max Welling has served as associate editor in chief of IEEE TPAMI from 2011-2015 and has been program chair for AISTATS (’09), NIPS (’13) and ECCV (’16) and general chair of NIPS in 2014. He received an NSF career grant in 2005 and is a recipient of the ECCV Koenderink Prize in 2010. Welling’s machine learning research labs are AMLAB, Qualcomm QUVA Lab and Bosch Delta Lab. Talk title: Artificial Intelligence per Kilowatt-hour Synopsis: The successes of deep learning are spectacular. But modern deep learning architectures with hundreds of layers and millions of parameters require an extraordinary amount of computation and data to train and execute. At the same time, more compute is moving to the edge. We predict that the next battle in AI is over how much intelligence can be squeezed out of every kilowatt-hour of energy. Max will discuss a number ideas to make progress on this problem, including compressing deep neural nets, computing with low bit precision and spiking neural networks. #aiattheturing
Artificial intelligence and society: In conversation with Jeff Sachs
 
01:37:41
Eighty years after Alan Turing launched the digital age, the revolutionary consequences continue to unfold. The changes are so relentless and powerful that they have given rise to competing utopian and dystopian narratives. According to one view, smart machines will usher in unparalleled productivity, prosperity, longevity, and leisure time; on the competing view, smart machines will crush workers, drive wealth to top 0.0001% and end privacy. About the speaker Professor Jeffrey Sachs will argue that both of these visions are coherent and theoretically possible, and that the outcome will depend not on the technology itself but on how we govern it. Rapid advances in AI and related technologies will replace workers, drive up the wages for some groups while driving down wages for others, and tend to increase inequalities of market earnings. Yet the benefits of the digital revolution can be broadly shared if we act with foresight, ethics, and appropriate strategies regarding taxation, intellectual property, education, and governance. #TuringLectures
Making machines that see: Geometry, Uncertainty and Deep Learning
 
45:40
Professor Cipolla has been a Professor of Information Engineering at the University of Cambridge since 2000. Previously he worked as a Toshiba Fellow and engineer at the Toshiba Corporation Research and Development Centre in Kawasaki, Japan, and was awarded a D.Phil. (Computer Vision) from the University of Oxford in 1991. Roberto’s research interests are in computer vision and robotics and include the recovery of motion and 3D shape of visible surfaces from image sequences; object detection and recognition; novel man-machine interfaces using hand, face and body gestures; real-time visual tracking for localisation and robot guidance; applications of computer vision in mobile phones, visual inspection and image-retrieval and video search. Talk title: Making machines that see: Geometry, Uncertainty and Deep Learning Synopsis: The last decade has seen a revolution in the theory and application of computer vision and machine learning. I will begin with a brief review of some of the fundamentals with a few examples of the reconstruction, registration and recognition of three-dimensional objects and their translation into novel commercial applications. I will then introduce some recent results from real-time deep learning systems that exploit geometry and compute model uncertainty. Understanding what a model does not know is a critical part of safe machine learning systems. New tools, such as Bayesian deep learning, provide a framework for understanding uncertainty in deep learning models, aiding interpretability and safety of such systems. Additionally, knowledge of geometry is an important consideration in designing effective algorithms. In particular, we will explore the use of geometry to help design networks that can be trained with unlabelled data for stereo and for human body pose and shape recovery. #aiattheturing
Professor Helen Margetts: "The Data Science of Politics"
 
01:02:05
The Turing Lectures: Social Science and Ethics - Professor Helen Margetts, Director, Oxford Internet Institute, University of Oxford: "The Data Science of Politics" Click the below timestamps to navigate the video. 00:00:07 Introduction by Professor Andrew Blake, Director, The Alan Turing Institute 00:01:40 Professor Helen Margetts, Director, Oxford Internet Institute, University of Oxford: "The Data Science of Politics" 00:50:01 Q&A The excitement of Data Science brings the need to consider the ethics associated with the information age. Likewise a revolution in political science is taking place where the internet, social media and real time electronic monitoring has brought about increased mobilisation of political movements. In addition the generation of huge amounts of data from such processes presents on the one hand opportunities to analyse and indeed predict political volatility, and on the other ethical and technical challenges which will be explored by two of the foremost philosophers and political scientists. The Alan Turing Institute is the UK's National Institute for Data Science. The Institute’s mission is to: undertake data science research at the intersection of computer science, mathematics, statistics and systems engineering; provide technically informed advice to policy makers on the wider implications of algorithms; enable researchers from industry and academia to work together to undertake research with practical applications; and act as a magnet for leaders in academia and industry from around the world to engage with the UK in data science and its applications. The Institute is headquartered at The British Library, at the heart of London’s knowledge quarter, and brings together leaders in advanced mathematics and computing science from the five founding universities and other partners. Its work is expected to encompass a wide range of scientific disciplines and be relevant to a large number of business sectors. For more information, please visit: https://turing.ac.uk #TuringLectures
Professor Stephen Roberts, University of Oxford
 
25:40
Bio Stephen Roberts is the RAEng/Man Professor of Machine Learning at the University of Oxford. Stephen is a Fellow of the Royal Academy of Engineering, the Royal Statistical Society, the IET and the Institute of Physics. Stephen is Director of the Oxford-Man Institute of Quantitative Finance and Director of the Oxford Centre for Doctoral Training in Autonomous Intelligent Machines and Systems (AIMS). Research Stephen’s interests lie in methods for machine learning & data analysis in complex problems, especially those in which noise and uncertainty abound. His current major interests include the application of machine learning to huge astrophysical data sets (for discovering exo-planets, pulsars and cosmological models), biodiversity monitoring (for detecting changes in ecology and spread of disease), smart networks (for reducing energy consumption and impact), sensor networks (to better acquire and model complex events) and finance (to provide time series and point process models and aggregate large numbers of information streams). #TuringShortTalks
Gamechangers for diversity in STEM
 
06:16
In September we hosted ‘Gamechangers for diversity in STEM’ – a two day hackathon bringing together 40 highly committed individuals from across the scientific community to tackle the problem of lack of diversity in science, technology, engineering and mathematics (STEM). It was an energetic and inspiring two days culminating in the development of nine innovative projects offering practical and sustainable long term solutions. This is a video of some of our participants and fantastic keynote speakers discussing their personal experiences and how the challenges can be overcome. You can also find out more about what happened at the event by reading the report: https://bit.ly/2FmAUJQ and listening to the podcast: https://bit.ly/2JRDNR8. Some of our participants have also written blogs which you can read on the Gamechangers website: https://www.stemgamechangers.github.io.
Introduction to machine learning: Professor Zoubin Ghahramani, Cambridge University
 
02:44:12
Zoubin Ghahramani FRS is Professor of Information Engineering at the University of Cambridge, where he leads the Machine Learning Group, and the Cambridge University Liaison Director of the Alan Turing Institute, the UK’s national institute for Data Science. He is also the Deputy Academic Director of the Leverhulme Centre for the Future of Intelligence, and a Fellow of St John’s College Cambridge. He has worked and studied at the University of Pennsylvania, MIT, the University of Toronto, the Gatsby Unit at UCL, and CMU. He is co-founder of Geometric Intelligence and advises a number of AI and machine learning companies. He has served as programme and general chair of the leading international conferences in machine learning: AISTATS, ICML, and NIPS. In 2015 he was elected a Fellow of the Royal Society. RESEARCH Zoubin’s current research interests include statistical machine learning, Bayesian nonparametrics, scalable inference, deep learning, probabilistic programming, Bayesian optimisation, and automating data science. His Automatic Statistician project aims to automate the exploratory analysis and modelling of data, discovering good models for data and generating a human-interpretable natural language summary of the analysis. He and his group have also worked on automating inference (though probabilistic programming) and on automating the allocation of computational resources. More information can be found at http://mlg.eng.cam.ac.uk/. #datascienceclasses
Professor Gareth Roberts: "New challenges in Computational Statistics"
 
01:03:49
The Turing Lectures: Statistics - Professor Gareth Roberts, University of Warwick “New challenges in Computational Statistics” Click the below timestamps to navigate the video. 00:00:09 Welcome by Professor Patrick Wolfe 00:01:44 Introduction by Professor Sofia Olhede 00:03:23 Professor Gareth Roberts, University of Warwick “New challenges in Computational Statistics” 00:59:59 Q&A The second set of Turing Lectures focuses on Statistical Science and we have two of the world’s leading statistical innovators giving two lectures on the new challenges in computational statistics and its application in life sciences. We will delve into the mysteries of the operation and control of the living cell, seeking to make sense of data obtained from ingenious experiments. Contemporary statistical models required for such complex data is presenting phenomenal challenges to existing algorithms and these talks will present advances being made in this area of Statistical Science. For more information, please visit: https://turing.ac.uk The Alan Turing Institute is the UK's National Institute for Data Science. The Institute’s mission is to: undertake data science research at the intersection of computer science, mathematics, statistics and systems engineering; provide technically informed advice to policy makers on the wider implications of algorithms; enable researchers from industry and academia to work together to undertake research with practical applications; and act as a magnet for leaders in academia and industry from around the world to engage with the UK in data science and its applications. The Institute is headquartered at The British Library, at the heart of London’s knowledge quarter, and brings together leaders in advanced mathematics and computing science from the five founding universities and other partners. Its work is expected to encompass a wide range of scientific disciplines and be relevant to a large number of business sectors. For more information, please visit: https://turing.ac.uk #TuringLectures
Data science applications: Urban data science - Professor Cecilia Mascolo, University of Cambridge
 
47:21
Bio Cecilia Mascolo is Full Professor of Mobile Systems in the Computer Laboratory, University of Cambridge, UK. Prior joining Cambridge in 2008, she has been a faculty member in the Department of Computer Science at University College London. Her research interests are in human mobility modelling, mobile and sensor systems and networking and spatio-temporal data analysis. Research In the Alan Turing Institute, Professor Mascolo hopes to research on aspects related related to interpretation and inference of mobile and wearable sensor data efficiently on devices, on how to make sense of this kind of data in ways which respect its fine grained spatial and temporal granularity. She hopes then to look at how to use the interpretation to improve systems and interventions for users. Her research is highly interdisciplinary and applications of this research will span various other disciplines like, to name a few, urban data science, mobile health and organization analytics. #datascienceclasses
Turing Lecture: Professor Jon Crowcroft, Cambridge University
 
58:32
Jon Crowcroft has been the Marconi Professor of Communications Systems in the Computer Laboratory since October 2001. He has worked in the area of Internet support for multimedia communications for over 30 years. Three main topics of interest have been scalable multicast routing, practical approaches to traffic management, and the design of deployable end-to-end protocols. Current active research areas are Opportunistic Communications, Social Networks, and techniques and algorithms to scale infrastructure-free mobile systems. He leans towards a “build and learn” paradigm for research. He graduated in Physics from Trinity College, University of Cambridge in 1979, gained an MSc in Computing in 1981 and PhD in 1993, both from UCL. He is a Fellow the Royal Society, a Fellow of the ACM, a Fellow of the British Computer Society, a Fellow of the IET and the Royal Academy of Engineering and a Fellow of the IEEE. He likes teaching, and has published a few books based on learning materials. Research Computing Systems at scale are the basis for much of the excitement over Data Science, but there are many challenges to continue to address ever larger amounts of data, but also to provide tools and techniques implemented in robust software, that are usable by statisticians and machine learning experts without themselves having to become experts in cloud computing. This vision of distributed computing only really works for “embarrassingly parallel” scenarios. The challenge for the research community is to build systems to support more complex models and algorithms that do not so easily partition into independent chunks; and to give answers in near real-time on a given size data centre efficiently. Users want to integrate different tools (for example, R on Spark); don’t want to have to program for fault tolerance, yet as their tasks & data grow, will have to manage this; meanwhile, data science workloads don’t resemble traditional computer science batch or single-user interactive models. These systems put novel requirements on data centre networking operating systems, storage systems, databases, and programming languages and runtimes. As a communications systems researcher for 30 years, I am also interested in specific areas that involve networks, whether as technologies (the Internet, Transportation etc), or as observed phenomena (Social Media), or in abstract (graphs). #TuringLectures
From Z3 to Lean, Efficient Verification - Dr Leonardo de Moura
 
26:02
https://www.turing-gateway.cam.ac.uk/sites/default/files/asset/doc/1707/from_z3_to_lean.pdf #TuringSeminars
A gentle introduction to network science: Dr Renaud Lambiotte, University of Oxford
 
01:40:43
The language of networks and graphs has become a ubiquitous tool to analyse systems in domains ranging from biology to physics and from computer science to sociology. Renaud will present important properties observed in real-life networked systems, as well as tools to understand and model their structures. #datascienceclasses
The unintended consequences of GDPR: Michael Ross, DynamicAction
 
16:30
Michael has spent the last 25 years at the intersection of the digital and data worlds. He is currently the Co-founder and Chief Scientist of DynamicAction which is a leader in big data analytics and AI for retail. He was previously the founder and CEO of figleaves.com, and started his career at McKinsey consulting in the early days of the internet. He is a non-executive director of Sainsbury’s Bank and N.Brown plc, and is on the Commercial Development Board of the Turing. Event information: The General Data Protection Regulation (GDPR) will come into force in the European Union in May 2018. The regulation is designed to strengthen the European data protection regime for personal data of EU residents which is processed within the EU and outside it. Notably, the GDPR also addresses the use of automated algorithmic decision-making and profiling in processing personal data, raising questions about the extent to which data subjects have rights to meaningful information about the logic involved, to obtain human intervention, and to contest decisions made by solely automated systems. By addressing a wide array of concepts, including fairness, transparency, privacy, consent, and interpretability, the GDPR is set to reshape the relationship between governments, corporations, and the individuals whose personal data they process. Since organisations found not to be in compliance with the regulation will face serious penalties (up to 4% of global revenue), there is great interest in exactly what the GDPR does and does not require, as well as how it will be interpreted after implementation. This day-long, expert-led workshop will explain the purposes and provisions of the GDPR, and explore what next steps might be for the regulation of artificial intelligence.
Jane Hillston (DDMCS@Turing): Moment analysis, model reduction and London bike sharing
 
24:48
Complex models in all areas of science and engineering, and in the social sciences, must be reduced to a relatively small number of variables for practical computation and accurate prediction. In general, it is difficult to identify and parameterize the crucial features that must be incorporated into the model, but powerful statistical approaches are becoming available based on analysis of large volumes of data. This is bringing fundamental change in the way we think about models. A new modelling paradigm is emerging based on the combination of statistical inference, high-throughput computation and physical laws, yet the mathematical foundations for combining these methods are still in their infancy. The purpose of this workshop is to bring together a diverse group of mathematicians and computational scientists to explore new ways of incorporating data analysis into complex systems modelling. Application topics to be discussed include methods for collective dynamics (flocking, schooling and pedestrian models), molecular modelling, cell biology, and fluid dynamics. The programme on each day will consist primarily of invited talks addressing diverse perspectives on data-driven modelling of complex systems. There will also be a poster session, and many opportunities for informal interaction and discussion.
Recurrent Neural Networks and Models of Computation - Edward Grefenstette, DeepMind
 
39:11
This talk presents an analysis of various recurrent neural network architectures in terms of traditional models of computation. It makes the case for simpler recurrent architectures being closer to finite state automata, and argues that memory-enhanced architectures support better algorithmic efficiency, even in problems which are describable as regular languages. Logic has proved in the last decades a powerful tool in understanding complex systems. It is instrumental in the development of formal methods, which are mathematically based techniques obsessing on hard guarantees. Learning is a pervasive paradigm which has seen tremendous success recently. The use of statistical approaches yields practical solutions to problems which yesterday seemed out of reach. These two mindsets should not be kept apart, and many efforts have been made recently to combine the formal reasoning offered by logic and the power of learning. The goal of this workshop is to bring together expertise from various areas to try and understand the opportunities offered by combining logic and learning. The programme has four axes, starting from a theoretical standpoint and going to a more practical one: logic and automata, verification, programming languages, and neural networks. Registration for this event is now closed, but if you would like to be added to the waiting list please email: [email protected]
Introduction to Data Ethics - Brent Mittelstadt
 
01:27:00
Dr. Brent Mittelstadt is a Research Fellow at the Alan Turing Institute and University College London. His research addresses the ethics of algorithms, machine learning, artificial intelligence and data analytics (‘Big Data’). Over the past five years his focus has broadly been on the ethics and governance of emerging information technologies, including a special interest in medical applications. Research Dr. Mittelstadt;'s research focuses on ethical auditing of algorithms, including the development of standards and methods to ensure fairness, accountability, transparency, interpretability and group privacy in complex algorithmic systems. His work addresses norms and methods for prevention and systematic identification of discriminatory and ethically problematic outcomes in decisions made by algorithmic and artificially intelligent systems. A recent paper on the legally dubious right to explanation and the lack of meaningful and accountability and transparency mechanisms for automated decision-making in the General Data Protection Regulation, co-authored with Dr. Sandra Wachter and Prof. Luciano Floridi, highlights the pressing need for work in these areas. #datascienceclasses
Introduction: Data Sciences for Climate and Environment
 
09:04
This is a short introduction to the Data Sciences for Climate and Environment by Professors Richard Smith and Mark Girolami. You can view the full event here: https://www.youtube.com/playlist?list=PLuD_SqLtxSdUVT_2SSPzZSC__kAxpkm8w About the event Collectively, we are modelling and monitoring our planet better than we have ever done in our history, as a result of sustained efforts from the climate modelling community and space agencies and the private sector worldwide. Climate and weather models can now be run at finer spatial resolutions (10km or better), therefore enabling more realistic simulations of smaller and smaller scale processes (i.e. tropical cyclones in the atmosphere or eddies in the ocean) that can have severe impacts on our planet. At the same time there is a rapid growth in the number of satellites orbiting the Earth (221 launched in 2015, around 5000 in total) with a significant fraction of these satellites dedicated to Earth Observation using a large variety of sensors working at different electromagnetic frequencies (optical, radar, infrared, etc.). Our ability to store, process and share efficiently the vast amounts of data that are produced (~Pb yearly) by the modelling and remote sensing communities is a pre-requisite for the good functioning of these often publicly funded large programmes. In this one-day workshop our speakers will present on how the new tools developed in data sciences can be applied to questions relating to climate and the environment to help us address the great challenges that our society is facing in a rapidly changing planet. Our event will be structured around five keynote speakers highlighting five separate topics described below and followed by a panel dialogue between our experts and the audience on the topic of Data Sciences for the Climate and the Environment.
Fellow Short Talks: Dr Peter Richtarik, Edinburgh University
 
30:59
Peter Richtarik is a Reader in the School of Mathematics at the University of Edinburgh, and is the Head of a Big Data Optimization Lab. He received his PhD from Cornell University in 2007, and currently holds an EPSRC Early Career Fellowship in Mathematical Sciences. RESEARCH My main research focus is the development of new optimization algorithms and theory. In particular, much of my recent work is in the emerging field of big data optimization, with applications in machine learning in general and empirical risk minimization in particular. For big data optimization problems, traditional methods are no longer suitable, and hence there is need to develop new algorithmic paradigms. An important role in this respect is played by randomized algorithms of various flavors, including randomized coordinate descent, stochastic gradient descent, randomized subspace descent and randomized quasi-Newton methods. Parallel and distributed variants are of particular importance. #TuringShortTalks