Events

Forthcoming events

This page lists C2D3 events, University events, as well as related external conferences and events of interest to our members.

Trustworthy AI
C2D3 event
Monday, 26 June 2023, 9.30am to 5.30pm

Principal Partner : Kavli Centre for Ethics, Science and the Public

Other supporting partners : AI@Cam, Bitfount, Centre for Human-Inspired Artificial Intelligence (CHIA) and Cambridge Global Consulting Ltd (CGC). 

knowledge graph
External event
Friday, 16 June 2023, 10.00am to 6.00pm

The 2nd Symposium of The Turing Interest Group on Knowledge Graphs will feature the keynotes of Lora Aroyo and Chris Welty (Google Research), and Artur Garcez (City, University of London). There will also be a special session on Neurosymbolic AI, short presentations by the group members and networking activities. The symposium is sponsored by Telicent, The Alan Turing Institute and City, University of London.

Poster
University of Cambridge event
Saturday, 17 June 2023, 9.00am to Sunday, 18 June 2023, 5.00pm

The Wellcome PhD programme on Mathematical Genomics and Medicine invites alumni of the University of Cambridge to its Alumni Event.

Registration

Participants register here.

Limited financial support may be available to assist alumni with travel/accommodation. Please use this separate form to request financial assistance.

medicine blue green thumbnail
External event
Tuesday, 27 June 2023, 3.00pm to 4.00pm

The Turing-Roche strategic partnership is a collaboration in advanced analytics between Roche and the Turing, focused on enabling the transformative benefits of personalised healthcare to become a reality for patients around the world. The Turing-Roche knowledge share series brings together members of Roche and Turing's networks to showcase partnership research and share

DS in UK secondaries marketing graphic
University of Cambridge event
Thursday, 29 June 2023, 10.00am to 4.00pm

The workshop will explore how data science might build on the rapid expansion of online learning platforms in UK schools to improve the language development and writing skills of students in the humanities and languages in the UK state school sector. Academics, together with teachers and the educational industry will explore how to frame a partnership for data sharing and educational AI applications in order to support language and literacy and improve outcomes in the UK state school sector.

The event is supported by the Alan Turing Institute.

UBAI 2023
C2D3 event
Friday, 30 June 2023, 9.00am to 6.00pm

This in-person conference aims to start a conversation around the role of AI in scientific understanding, and whether breakthroughs in the intersection of AI and biology are pushing us to revise our notion of scientific understanding. 

It is open to researchers along with stakeholders in industry and society.

 

Full, up-to-date details and participant registration can be found on the UBAI 2023 website.

 

poster
C2D3 event
Thursday, 21 September 2023, 9.00am to 5.00pm

The University of Cambridge and EMBL-EBI are hosting a joint one-day computational biology symposium to explore collaborative research and future opportunities between the two institutions. 

Forthcoming talks

A collation of interesting data science talks from across the University.

Game theory, distributional reinforcement learning, control and verification

Wednesday, 7 June 2023, 12.00pm to 1.30pm
Speaker: Prof. Alessandro Abate, Dr. Licio Romao, Dr. Yulong Gao and Dr. Jiarui Gan. University of Oxford
Venue: Cambridge University Engineering Department, CBL Seminar room BE4-38.

This week, the MLG looks forward to welcoming four guest speakers from Oxford.

*Talk 1:*

_Title:_ Formal Synthesis with Neural Templates
Speaker: Prof. Alessandro Abate (Dept. Computer Science, Univ. of Oxford, UK)

_Abstract:_ I shall present recent work on CEGIS, a "counterexample-guided inductive synthesis'' framework for sound synthesis tasks that are relevant for dynamical models, control problems, and software programs. The inductive synthesis framework comprises the interaction of two components, a learner and a verifier. The learner trains a neural template on finite samples. The verifier soundly validates the candidates trained by the learner, by means of calls to a SAT-modulo-theory solver. Whenever the candidate is not valid, SMT-generated counter-examples are passed to the learner for further training.

_Bio:_ Alessandro Abate is Professor of Verification and Control in the Department of Computer Science at the University of Oxford, where he is also Deputy Head of Department. Earlier, he did research at Stanford University and at SRI International, and was an Assistant Professor at the Delft Center for Systems and Control, TU Delft. He received an MS/PhD from the University of Padova and UC Berkeley. His research interests lie on the formal verification and control of stochastic hybrid systems, and in their applications in cyber-physical systems, particularly involving safety criticality and energy. He blends in techniques from machine learning and AI, such as Bayesian inference, reinforcement learning, and game theory.

*Talk 2:*

_Title:_ Policy synthesis with guarantees

_Speaker:_ Dr. Licio Romao (Dept. Computer Science, Univ. of Oxford, UK)

_Abstract:_ In this talk, I will present two techniques to perform feedback policy synthesis with guarantees. First, I will introduce a new concept of RL robustness and show how to obtain the best robust policy within a class of sub-optimal solutions by leveraging lexicographic optimisation. The proposed notion of robustness is motivated by the fact that, at deployment, the state of the system may not be precisely known due to measurement errors. In the second part of the talk, I will present a new technique to derive abstractions of stochastic dynamical systems. Our methodology is agnostic to the probability measure that generates the noise and leads to an interval Markov Decision Process (iMDP) representation of the original dynamics; the interval transition probability contains, with high probability, the true transition probability between states of the abstraction. The PAC guarantees of the proposed framework are obtained due to a non-trivial connection with the scenario approach theory, a technique that has had tremendous success within the control community.

_Bio:_ Licio Romao is a postdoctoral research assistant in the Department of Computer Science at the University of Oxford. He obtained his PhD in August 2021 from the Department of Engineering Science, and MSc and BSc from the University of Campinas (UNICAMP) and the Federal University of Campina Grande (UFCG), respectively. His PhD thesis was awarded the Institute of Engineering Technology’s (IET) Control and Automation Dissertation Prize 2021. His research combines techniques from formal verification, control theory, applied mathematics, and machine learning to enable the design of safer and more reliable feedback systems.
_Relevant papers:_
· D. Jarne, L. Romao, L. Hammond, M. Mazo Jr, A. Abate. Observational Robustness and Invariances in Reinforcement Learning via Lexicographic Objectives. 2023. Link: https://licioromao.com/assets/papers/JRHMA23.pdf.
· T. Badings, L. Romao, A. Abate, D. Parker, H. Poonwala, M. Stoelinga, N. Jensen. Robust Control for Dynamical Systems with Non-Gaussian via Formal Abstractions. Journal of Artificial Inteligence Research. 2023. Link: https://licioromao.com/assets/papers/BRAPPSJ23.pdf.
· T. Badings, L. Romao, A. Abate, N. Jensen. Probabilities are not enough: formal controller synthesis for stochastic dynamical systems with epistemic uncertainty. AAAI Conference On Artificial Intelligence , 2023. Link: https://licioromao.com/assets/papers/BRAJ23a.pdf.

*Talk 3:*

_Title:_ Policy Evaluation in Distributional LQR

_Speaker:_ Dr. Yulong Gao (Dept. Computer Science, Univ. of Oxford, UK)

_Abstract:_ Distributional reinforcement learning (DRL) enhances the understanding of the effects of the randomness in the environment by letting agents learn the distribution of a random return, rather than its expected value as in standard RL. At the same time, a main challenge in DRL is that policy evaluation in DRL typically relies on the representation of the return distribution, which needs to be carefully designed. In this talk, I will discuss a special class of DRL problems that rely on discounted linear quadratic regulator (LQR) for control, advocating for a new distributional approach to LQR, which we call distributional LQR. Specifically, we provide a closed-form expression of the distribution of the random return which, remarkably, is applicable to all exoge- nous disturbances on the dynamics, as long as they are independent and identically distributed (i.i.d.). While the proposed exact return distribution consists of infinitely many random variables, we show that this distribution can be approximated by a finite number of random variables, and the associated approximation error can be analytically bounded under mild assumptions. Using the approximate return distribution, we propose a zeroth-order policy gradient algorithm for risk-averse LQR using the Conditional Value at Risk (CVaR) as a measure of risk. Numerical experiments are provided to illustrate our theoretical results. (https://arxiv.org/abs/2303.13657)

_Bio:_ Yulong Gao is a postdoctoral researcher at the Department of Computer Science, University of Oxford. He received the joint Ph.D. degree in Electrical Engineering in 2021 from KTH Royal Institute of Technology, Sweden, and Nanyang Technological University, Singapore. Before moving to Oxford, he was a Researcher at KTH from 2021 to 2022. He was the receipt of the VR International Postdoc Grant from Swedish Research Council. His research interests include automatic verification, stochastic control and model predictive control with application to safety-critical systems.

*Talk 4:*

_Title:_ Sequential information and mechanism design

_Speaker:_ Dr. Jiarui Gan (Dept. Computer Science, Univ. of Oxford, UK)

_Abstract:_ Many problems in game theory involve reasoning between multiple parties with asymmetric access to information. This broad class of problems lead to many research questions about information and mechanism design, with broad-ranging applications from governance and public administration to e-commerce and financial services. In particular, there has been a recent surge of interest in exploring the more generalized sequential versions of these problems, where players interact over multiple time steps in a changing environment. In this talk, I will present a framework of sequential principal-agent problems that is capable of modeling a wide range of information and mechanism design problems. I will talk about our recent algorithmic results on the computation and learning of optimal decision-making in this framework.

_Bio:_ Jiarui Gan is a Departmental Lecturer at the Computer Science Department, University of Oxford, working in the Artificial Intelligence & Machine Learning research theme. Before this he was a postdoctoral researcher at Max Planck Institute for Software Systems, and he obtained his PhD from Oxford. Jiarui is broadly interested in algorithmic problems in game theory. His current focus is on sequential information and mechanism design problems. His recent work has been selected for an Outstanding Paper Honorable Mention at the AAAI'22 conference.

Segment Anything in Medical Images

Wednesday, 7 June 2023, 2.00pm to 3.00pm
Speaker: Dr Bo Wang, University of Toronto and Vector Institute of Artificial Intelligence
Venue: Virtual (see abstract for Zoom link)

Medical imaging plays an indispensable role in clinical practice. Accurate and efficient medical image segmentation provides a means of delineating regions of interest and quantifying various clinical metrics. However, building customized segmentation models for each medical imaging task can be a daunting and time-consuming process, limiting the widespread adoption in clinical practice. In this talk, I will introduce MedSAM, a segmentation foundation model that enables universal segmentation across a wide range of medical imaging tasks and modalities. MedSAM achieved remarkable improvements in 30 segmentation tasks, surpassing the existing segmentation foundation model by a large margin. MedSAM also demonstrated zero-shot and few-shot capabilities to segment unseen tumor types and adapt to new imaging modalities with minimal effort. Our results validate the versatility of MedSAM compared to existing customized segmentation models, emphasizing its potential to transform medical image segmentation and enhance clinical practice. This work underscores the significance of creating adaptable and efficient segmentation tools that can meet the growing demands of personalized healthcare and contribute to the ongoing progress in medical imaging analysis.

This seminar will be held online via ZOOM.
*Join Zoom Meeting:* https://maths-cam-ac-uk.zoom.us/j/93331132587?pwd=MlpReFY3MVpyVThlSi85Tm...

Untangling genome assembly graphs with graph neural networks

Wednesday, 7 June 2023, 5.00pm to 6.00pm
Speaker: Lovro Vrcek, Genome Institute of Singapore, A*STAR
Venue: Seminar Room FW26, Computer Laboratory, William Gates Building

With the emergence of PacBio HiFi and ultra-long ONT reads, the efforts to assemble genomes of various species in a de novo manner have significantly increased. This is manifested in projects including the T2T consortium project, the Human Pangenome Project and the Vertebrate Genome Project, which strive to assemble a large number of genomes with contemporary tools and data. However, even with all the recent advances in sequencing technologies, manual curation of the assembly genomes is still necessary. At the same time, most de novo assembly tools rely on graph-simplification heuristics, which have remained largely unchanged in recent years. Moreover, heuristics parameters have been hand-crafted using several genomes for which a high-quality reference was available during the time of development.

We implemented an entirely novel approach for resolving assembly graphs into genomes, one based on graph neural networks. We evaluated our method on different types of reads and with initial assembly graphs produced in a different way, comparing it against state-of-the-art de novo assemblers used in the field. Moreover, the preliminary results indicate the modularity and the adaptability of our approach, which should generalize better with every new genome assembly released, requiring minimal adjustments to the existing pipeline.

Statistics Clinic Easter 2023 IV

Wednesday, 7 June 2023, 5.30pm to 7.00pm
Speaker: Speaker to be confirmed
Venue: MR14

This free 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.

If you would like to participate, please sign up as we will not be able to offer a consultation otherwise. Please sign up through the following form: https://forms.gle/x7jvqZo6VigwGZm97. Sign-up is possible from Jun 1 midday until Jun 5 midday or until we reach full capacity, whichever is earlier. If you successfully signed up, we will confirm your appointment by Jun 7 midday.

Modeling crossmodal attention in humanoid robots for HRI in complex social scenarios

Thursday, 8 June 2023, 12.00pm to 1.00pm
Speaker: Di Fu, University of Hamburg
Venue: Zoom: https://cl-cam-ac-uk.zoom.us/j/96901178502?pwd=ODJOZ0pDQnVaeGNleHpIUDRNTUF6QT09

Abstract: Due to the aging population and life digitalization, humanoid robots could be seen as potential assistants to accompany senior citizens, support remote work, and improve individuals' mental or physical health. It is essential for robots to become more socialized by processing multiple social cues in a complex real-world environment. Thus, social cues integration, crossmodal attention, and conflict resolution are crucial for humanoid robots to implement social interaction. I will present a series of studies to explore the following questions: 1) how a humanoid robot could perform human-like attentional behaviors in complex environments with crossmodal conflicts; 2) how a humanoid robot interacts with humans by integrating different social cues and exhibiting different personalities; 3) how gaze and facial expressions of a humanoid robot impact on human-human-robot collaboration. Our experiments explored human crossmodal attention mechanisms with high ecological validity and proved that a humanoid robot could replicate human-like responses. Our interdisciplinary work provides important insights into how crossmodal attention can be modeled in robots and introduces future research directions for HRI.

Biography: Dr. Di Fu completed her doctoral training in human-robot interaction at the Department of Informatics, the University of Hamburg (UHH), with Prof. Stefan Wermter from 2017 to 2020, and in cognitive neuroscience at the Institute of Psychology, Chinese Academy of Sciences (CAS) with Prof. Xun Liu from 2014 to 2020. She is currently a Postdoctoral Researcher at the University of Hamburg, with research interests in audiovisual crossmodal attention, social robotics, and human-robot interaction. She had been honored as an outstanding graduate of CAS and an outstanding doctoral graduate of Beijing. She has been awarded the Kavli Summer Institute in Cognitive Neuroscience fellowship, the International postdoctoral exchange fellowship, the CAS-DAAD joint doctoral student fellowship, and the Chinese National Academic Scholarship. Her work has been published in the International Journal of Social Robotics, Public Administration Review, IEEE IROS, IEEE RO-MAN, IEEE IJCNN, etc. She also serves as a committee member of the Chinese Association for Psychological & Brain Sciences and the Chinese German Association for Biology and Medicine.