Menu

Home / Events / C2D3 Virtual Symposium 2020 / Programme and Abstracts: C2D3 Virtual Symposium 2020

Programme and Abstracts: C2D3 Virtual Symposium 2020

Back to: 
C2D3 Virtual Symposium 2020

Programme: C2D3 Virtual Symposium 2020 Research Rendezvous

Each talk will be accompanied by a live meet the speaker Q&A session

E-Posters

E-Posters will be available to view online from the day before the symposium and during the symposium. Voting will end at 23:59 (21 Oct) with the results announced on the 22 Oct via email and social media.

Programme

 

Abstracts: C2D3 Virtual Symposium 2020 Research Rendezvous

Academic Keynote Speaker: Prof Mihaela van der Schaar

John Humphrey Plummer Professor of Machine Learning, Artificial Intelligence and Medicine at the University of Cambridge, a Fellow at The Alan Turing Institute in London, and a Chancellor’s Professor at UCLA

Title: AutoML: powering the new human-machine learning ecosystem

Abstract: In this keynote session, I will explain the unique characteristics of healthcare that make it a challenging but extremely promising domain in which to apply AutoML. I will give an overview of several novel approaches we have developed to tackle problems as complex and diverse as AutoML for survival analysis, causal inference, and dynamic forecasting from time-series data. I will also highlight medical AutoML frameworks used in real-world contexts, including predictive tools deployed in response to the COVID-19 pandemic.

Biography: Mihaela was elected IEEE Fellow in 2009. She has received numerous awards, including the Oon Prize on Preventative Medicine from the University of Cambridge (2018), a National Science Foundation CAREER Award (2004), 3 IBM Faculty Awards, the IBM Exploratory Stream Analytics Innovation Award, the Philips Make a Difference Award and several best paper awards, including the IEEE Darlington Award.

Mihaela’s work has also led to 35 USA patents (many widely cited and adopted in standards) and 45+ contributions to international standards for which she received 3 International ISO (International Organization for Standardization) Awards.

In 2019, she was identified by National Endowment for Science, Technology and the Arts as the most-cited female AI researcher in the UK. She was also elected as a 2019 “Star in Computer Networking and Communications” by N²Women. Her research expertise spans signal and image processing, communication networks, network science, multimedia, game theory, distributed systems, machine learning and AI.

Mihaela’s current research focus is on machine learning, AI and operations research for healthcare and medicine. https://www.vanderschaar-lab.com/prof-mihaela-van-der-schaar/

 

Guest Keynote Speaker: Dr Orlando Machado

Chief Data Scientist, Aviva Quantum

Title: Data, AI, and connecting with customers 

Abstract: This talk will cover a number of ways in which the rapid growth of data sources, combined with advances in data science techniques, can help businesses build better customer experiences. Using examples from Aviva’s digital transformation, the talk will also highlight ways in which real-life commercial applications of technology require a different focus compared to work in an academic setting. 

Biography: Orlando Machado is Chief Data Scientist of Aviva Quantum, a global data science practice with over 500 members. Starting his career in academia, he has spent over 20 years helping organisations get value out of their data…long before ‘data scientist’ was crowned the ‘sexiest job of the 21st century’. 

In 2019, Orlando was ranked #1 'DataIQ' list of the 100 most influential people in data-driven business. Orlando is based in Aviva’s Digital Garage in London’s fashionable Hoxton Square. He holds a PhD in Statistics from the University of Warwick. Prior to joining Aviva in 2016, Orlando was: Chief Data Scientist at MoneySuperMarket, the UK's largest price comparison website; Head of Customer Insight at dunnhumby, a company at the cutting edge of data science for more than 20 years; Head of Analytics at Wunderman, one of the world's largest communications agencies. 

 

Dr Rosana Collepardo

Interdisciplinary Lecturer at the Departments of Chemistry and Genetics, and a Winton Advanced Research Fellow at the Department of Physics, University of Cambridge

Title: Multiscale modelling of genome organization

Abstract: Reading the genome is one thing, finding out how it functions is something else altogether. How the genome is organized in space and how this organization influences its function remains unclear. In this talk, l will introduce our new multiscale modelling toolkit designed to investigate unknown molecular mechanisms that dictate folding of the genome under varying conditions (e.g. epigenetic marks, protein binding, DNA breathing). I will explain how the structure of chromatin is much more fluid and irregular than originally proposed (Collepardo and Schlick, PNAS 2014; Sridhar and Collepardo, PNAS 2020) and how multiscale modelling can be used to reveal the molecular mechanisms behind epigenetic control of chromatin structure (Collepardo et al., JACS, 2015; Sridhar and Collepardo, PNAS 2020). Finally, I will talk about the new paradigm in the field of genome organization – DNA compartmentalization via liquid-liquid phase separation – and our newest work in that area (Espinosa, PNAS 2020).

Biography: Rosana Collepardo is an Interdisciplinary Lecturer at the Departments of Chemistry and Genetics, and a Winton Advanced Research Fellow at the Department of Physics, University of Cambridge. Rosana did her DPhil with David Manolopoulos at the PTCL in the University of Oxford, and trained as a postdoc with Daan Frenkel and David Wales here at the Department of Chemistry, with Modesto Orozco at the Barcelona Supercomputing Centre, and with Tamar Schlick at New York University.

Her group has been recently awarded a €1.49 million ERC starting grant to investigate genome organization with sub-nucleosome resolution via computational modelling.

Guided by experimental observations, physical and chemical information, advanced computer simulation techniques, and the tools of statistical mechanics, the Collepardo group develops theoretical and computational methods to elucidate the physical mechanisms that determine packaging of DNA inside Eukaryotic cells and the physical determinants for the spatiotemporal organization of molecules insides cells.

 

Dr Ajith Parlikad

Asset Management Group at the Institute for Manufacturing, University of Cambridge

Title: Digital Twins of Industrial Systems

Abstract: We are living in a digital era. It is evident that the practice of maintenance and asset management has the potential to be one of the biggest beneficiaries of this digital revolution. Concepts such as Industrial Internet of Things, Cyber Physical Systems and Digital Twins are now seen as major opportunities for companies in the manufacturing and infrastructure sectors to improve their products, processes and services. The linkage between the real world and the virtual world enabled by digital twins - supported by new data analytics and innovative machine learning techniques - allow optimised management of industrial systems. This session will explore this opportunity by focussing on our latest research on data-driven prognostics of asset failures, predictive maintenance, and asset fleet optimisation using Digital Twins.

Biography: Dr Ajith Kumar Parlikad is Reader in Asset Management at Cambridge University Engineering Department (CUED), where he leads a research team focused on examining how emerging data and digital technologies can be exploited to improve the whole-life performance of industrial and infrastructure systems. He holds an extensive research portfolio to address the fundamental technical challenges funded by the EPSRC, Innovate UK, EU H2020 and Industry. Dr Parlikad is the scientific secretary of the IFAC TC 5.1 Working Group on Advanced Maintenance Engineering, Services and Technology, and a member of the UK Digital Twin Hub.

 

Dr Mireia Crispin

Borysiewicz Fellow, Cancer Research UK - Cambridge Institute, University of Cambridge

Title: Augmented radiology: the challenges of image analysis and data integration

Abstract: Most of the work done in the area of data science for radiology has focused on the automation of tasks or processes that are typically done manually. However, the quantitative analysis of standard image data –for example, the CT scan of a cancer patient—can give us more insights into the complexity of the disease, going beyond the usual metrics.

In this talk I will discuss the question of quantifying tumour heterogeneity, an important biological concept that can drive the evolution of the disease. I will focus on two different approaches: a physics-driven subsegmentation of different types of tissue; and a data-driven exploration of image textures that could represent variations in the tumour microenvironment. We will see that often, in this type of patient-focused research, our work has to go beyond what is strictly data science in order to validate our findings and obtain rich datasets.

Biography: Mireia Crispin is a Research Fellow at Trinity College, University of Cambridge and Cancer Research UK, and holds a Borysiewicz Fellowship from the University of Cambridge. She is the co-lead of the Computational Group of the Mark Foundation Institute for Integrative Cancer Medicine, focusing on the development of machine learning models for cancer. Mireia also works at the intersection between science and policy as Director of the Healthcare Innovation programme of the Center for the Governance of Change at IE University (Madrid, Spain), which currently focuses on the integration of AI in European healthcare. She is an incoming member of the University of Cambridge Research Policy Committee, responsible for advising the General Board on strategic matters relating to the research activities of the University.

She previously worked at Memorial Sloan Kettering Cancer Center in New York, where she was also co-Chair of the New York Science and Education Policy Association. Mireia holds a PhD in Particle Physics from the University of Oxford, where she worked on data analysis methods for the search for dark matter at the Large Hadron Collider at CERN.

 

Ms Reema Patel

Head of Public Engagement, Ada Lovelace Institute

Title: Democratising dialogue on data and AI:

Abstract: Reema will make the case for why data and AI needs to be democratised, placing legitimacy at the heart of ensuring citizens can shape technology's direction and purpose. She will draw upon examples of how best to enable a broader public conversation about the challenges and issues data governance presents.

Biography: Reema Patel is Head of Public Engagement at the Ada Lovelace Institute, working to ensure that data and AI work for people and society. She is an experienced policy professional who has led citizen engagement and participation initiatives on complex and controversial policy areas in the UK, including the Royal Society of Arts (RSA) Citizens’ Economic Council, which successfully worked with and influenced the Bank of England’s public engagement strategy. Reema has consulted for a variety of international organisations, including the Danish Board of Technology Foundation and San Francisco-based technology social media start-up Nextdoor.com. She is a fellow of the RSA, founding trustee of a community-run library, and a local councillor.

 

Parviz Peiravi

Global CTO/Principal Engineer, Financial Services Industry Solutions, Solutions Architecture Development, Sales and Marketing Group, Intel

Title: Revolutionizing Data Collaboration with Trusted Multi-Party Federated Machine Learning

Abstract: We live in a world that’s becoming more data-driven every day. Organizations across a wide range of industries are using artificial intelligence (AI) and machine learning (ML) technologies to tap into complex data sets, unearth valuable insights and drive innovation. From healthcare to the financial sector and beyond, advanced data science models and big data projects are unlocking insights that can deliver everything from novel approaches to preventing and treating disease to highly effective financial fraud and money laundering detection.

But these projects aren’t without their challenges. Organizations looking to embark on data collaboration initiatives must overcome obstacles such as data ownership issues, compliance and a wave of data privacy requirements such as GDPR, Data sovereignty etc. In today’s data-filled world, ensuring privacy and security is paramount, and the measures to which organizations must go to achieve this can make collaborative data science difficult. A distributed machine learning method first introduced by Google about five years ago, Federated Machine Learning offers tremendous advantages when it comes to privately and securely enabling model training against large pools of data from multiple entities. It takes the opposite approach of the previous technique, meaning a Federated Machine Learning will bring aggregation to the data sources, rather than requiring all participating organizations to move their data sets to a centralized compute environment for aggregation. In this session we review architecture and technologies that enable trusted multi-party federated learning.

Biography: Parviz Peiravi is a Global CTO for Financial Services Industry Solutions with Intel Corporation responsible for Financial Services Industry solution architecture development. He is primarily responsible for designing and driving development of Artificial Intelligence, Big Data, Service Oriented/Microservices Architecture, Cloud, and IoT computing architectures in support of Intel’s focus areas within financial Services Industry. Parviz has been working with number of financial institutions worldwide in designing innovative solutions, drive the digital transformation, using the most relevant advanced technologies and performance engineering principles such as Hybrid Multi-Cloud infrastructure, multi-channel systems for retail banking, KYC/AML/fraud detection with federated Learning, Market surveillance, risk management systems, marketing campaign management, algorithmic trading.

In addition he has been working on high performance computing infrastructure and data grid for actuarial calculation and credit risk exposure using for HPC Clusters. Legacy system migration to private, hybrid and public cloud. He has numerous certifications in Enterprise Architecture Framework (TOGAF), SOA, ITIL, XML\Web Services, VMware, Openstack, Hadoop, Spark and Database design. His current focus is researching the application of Blockchain technologies, Advanced Analytics (Big Data, Machine Learning and Cognitive Computing), Federated Learning, and Internet of Things across financial services industry domains. He is member of Silicon Valley CTO Professionals, Linux Foundation, Blockchain Hyper Ledger, Cloud Computing Group E3C, Cloud Security Alliance (CSA), DMTF, and other organizations. Parviz has been with Intel 23+ years and holds a degree in Computer and Electrical Engineering and a recipient of Intel Achievement Award (IAA) and Intel Quality Award (IQA).

Publication: “The Business Value of Virtual Service Oriented Grids” by Enrique Castro-leon, Jackson He, Mark Chang and Parviz Peiravi, Intel Press (2008), ISBN 978-1934053102. He is also editor in chief of “Journey to Cloud” eMagazine published by Intel Corporation.

 

Prof Jennifer Gabrys

Chair in Media, Culture and Environment, Department of Sociology, University of Cambridge

Title: Citizen Data: From Monitoring to Justice 

Abstract: The evolving relationship between citizens and data is a fundamental issue of our time. It impacts social formation, cohesion and civil rights, since data has become the basis for innumerable social, political and economic processes and decisions. While data can contribute to original social insights, at the same time numerous concerns have arisen, ranging from the pervasive tracking and surveillance, to ownership monopolies that restrict access and control for data analysis, and production. In order to address these concerns, people are engaging in alternative practices of production, ownership and data analysis. Through these practices they are attempting to challenge dominant data regimes by becoming active in the creation of alternative practices and infrastructures. This presentation will consider the changing democratic engagements that are emerging through citizen data practices, especially as they relate to environmental monitoring, digital advocacy and social justice. 

Biography: Jennifer Gabrys is Professor of Media, Culture and Environment in the Department of Sociology at the University of Cambridge. She is Principal Investigator on the project AirKit, and she leads the Citizen Sense project, both funded by the European Research Council. In May 2020, she is beginning a new ERC-funded project, Smart Forests: Transforming Environments into Social-Political Technologies. She is the author of Program Earth: Environmental Sensing Technology and the Making of a Computational Planet (2016); and Digital Rubbish: A Natural History of Electronics (2011); and co-editor of Accumulation: The Material Politics of Plastic (Routledge, 2013). Her recent and in-progress books include How to Do Things with Sensors (2019), and Citizens of Worlds: Open-Air Toolkits for Environmental Struggle. Her work can be found at citizensense.net and jennifergabrys.net.

Intel

 

 

Symposium Principal Sponsor

CUP

Symposium Sponsor

 

In support of the event are three Cambridge University Press Gold open access journals focused on the cutting-edge applications of data science in engineering, in public policy and in the growing body of environment related modelling and analysis. Full details can be found here and in the pdf attachement below.

Data-Centric Engineering DCE is an open access journal run by an international team of distinguished experts. Editor-in-Chief: Mark Girolami University of Cambridge (& C2D3 Steering Committee) & The Alan Turing Institute, UK

Data & Policy A peer-reviewed, open access journal dedicated to the impact of data science on policy and governance. Editors include: Jon Crowcroft University of Cambridge & Alan Turing Institute, UK

Newly Launched Title: Environmental Data Science An interdisciplinary, open access journal dedicated to the potential of artificial intelligence and data science to enhance our understanding of the environment, and to address climate change. 

INT logo

 

 

Event financial support

Forthcoming talks

Achieving Consistent Low Latency for Wireless Real-Time Communications with the Shortest Control Loop

Thursday, 18 August 2022, 4.00pm to 5.00pm
Speaker: Zili Meng, Tsinghua Unversity
Venue: FW11 and https://cl-cam-ac-uk.zoom.us/j/97216272378?pwd=M2diTFhMTnppckJtNWhFVTBKK0REZz09

Real-time communication (RTC) applications like video conferencing or cloud gaming require consistent low latency to provide a seamless interactive experience. However, wireless networks including WiFi and cellular, albeit providing a satisfactory median latency, drastically degrade at the tail due to frequent and substantial wireless bandwidth fluctuations. We observe that the control loop for the sending rate of RTC applications is inflated when congestion happens at the wireless access point (AP), resulting in untimely rate adaption to wireless dynamics. Existing solutions, however, suffer from the inflated control loop and fail to quickly adapt to bandwidth fluctuations. In this paper, we propose Zhuge, a pure wireless AP based solution that reduces the control loop of RTC applications by separating congestion feedback from congested queues. We design a Fortune Teller to precisely estimate per-packet wireless latency upon its arrival at the wireless AP. To make Zhuge deployable at scale, we also design a Feedback Updater that translates the estimated latency to comprehensible feedback messages for various protocols and immediately delivers them back to senders for rate adaption. Trace-driven and real-world evaluation shows that Zhuge reduces the ratio of large tail latency and RTC performance degradation by 17% to 95%.

Speaker Bio: Zili is a 3rd-year PhD student in Tsinghua University. His current research interest focuses on real-time video communications. He has published several papers in SIGCOMM / NSDI and received the Microsoft Research Asia PhD Fellowship, Gold Medal of SIGCOMM 2018 Student Research Competition, and two best paper awards.

BSU Seminar: "Genome-wide genetic models for association, heritability analyses and prediction"

Monday, 22 August 2022, 4.30pm to 5.30pm
Speaker: David Balding, Honorary Professor of Statistical Genetics at UCL Genetics Institute and University of Melbourne
Venue: Seminar Rooms 1 & 2, School of Clinical Medicine, Hills Road, Cambridge CB2 0SP

Although simultaneous analysis of genome-wide SNPs has been popular for over a decade, the problems posed by more SNPs than study participants (more parameters than data points), and correlations among the SNPs, have not been adequately overcome so that almost all published genome-wide analyses are suboptimal. While there has been much attention paid to the shape of prior distributions for SNP effect sizes, we argue that this attention is misplaced. We focus on what we call the "heritability model": a low-dimensional model for the expected heritability at each SNP, which is key to both individual-data and summary-statistic analyses. The 1-df uniform heritability model has been implicitly adopted in a wide range of analyses. Replacing it with better heritability models, using predictors based on allele frequency, linkage disequilibrium and functional annotations, leads to substantial improvements in estimates of heritability and selection parameters over traits, and over genome regions, as well as improvements in gene-based association testing and prediction. Key collaborators Doug Speed, Aarhus, Denmark and Melbourne PhD student Anubhav Kaphle.

Statistics Clinic Summer 2022 III

Wednesday, 31 August 2022, 5.30pm to 7.00pm
Speaker: Speaker to be confirmed
Venue: Venue to be confirmed

If you would like to participate, please fill in the following "form":https://forms.gle/b1UzrTNBig7hkr1e7. The deadline for signing up for a session is 12pm on Monday the 29th of August. Subject to availability of members of the Statistics Clinic team, we will confirm your in-person or remote appointment.

This event is open only to members of the University of Cambridge (and affiliated institutes). Please be aware that we are unable to offer consultations outside clinic hours.

Statistics Clinic Summer 2022 IV

Wednesday, 21 September 2022, 5.30pm to 7.00pm
Speaker: Speaker to be confirmed
Venue: Venue to be confirmed

Abstract not available

Title to be confirmed

Monday, 26 September 2022, 3.00pm to 4.00pm
Speaker: Christopher Yau, University of Manchester
Venue: CRUK CI Lecture Theatre

Abstract not available