Events

Forthcoming events

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

University of Cambridge event
Thursday, 2 February 2023, 5.45pm to 7.30pm

The AI Club is a new initiative to bring together the Biomedical AI and Machine Learning community in Cambridge, to discuss common themes and explore different topics and methodologies. We hope these thematic monthly sessions will open conversations to inform, inspire, and connect researchers at all levels, on topics within computational biology, AI and bioinformatics.

Turing Lecture February 2023 graphic
External event
Tuesday, 14 February 2023, 2.00pm to 3.30pm

Organiser: The Alan Turing Institute

In the first Turing Lecture for everyone aged 11+, wildlife filmmaker Tom Mustill reveals how a close encounter with a humpback whale inspired him to find out whether we could ever talk to these ocean giants.

We invite you to submerge yourself in Tom’s underwater world, hear about the amazing advances scientists are making in decoding animal communication, and be rumbled by the voices of our deep-sea orca-stra!

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. The deadline to register is 22nd February 2023.

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

Forthcoming talks

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

Statistics Clinic Lent 2023 II

Wednesday, 1 February 2023, 4.30pm to 6.00pm
Speaker: Speaker to be confirmed
Venue: MR5

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 link: https://forms.gle/Xbq3UiCPsiyjHJBQ6. Sign-up is possible from Jan 26 midday until Jan 30 midday or until we reach full capacity, whichever is earlier. If you successfully signed up, we will confirm your appointment by Feb 1 midday.

Narrative Summarization From Multiple Views

Thursday, 2 February 2023, 11.00am to 12.00pm
Speaker: Pinelopi (Nelly) Papalampidi, DeepMind
Venue: GR04, English Faculty Building, 9 West Road, Sidgwick Site

Although summarizing movies and TV shows comes naturally to humans, it is very challenging for machines. They have to combine different input sources (i.e., video, audio, subtitles), process long videos of 1-2 hours, and their transcripts, and learn from a handful of examples, since collecting and processing such videos is hard. Given the challenges of multimodal summarization, most prior work does not consider all facets of the computational problem at once but instead focuses on either processing multiple but short input sources or long text-only narratives.

In contrast, we aim at summarizing full-length movies and TV episodes while considering all input sources for creating video trailers and textual summaries. For trailer creation, we propose an algorithm for selecting trailer moments in movies based on interpretable criteria such as the narrative importance and sentiment intensity of events. We further demonstrate how we can convert our algorithm into an interactive tool for trailer creation with a human in the loop. Next, for producing textual summaries from full-length TV episodes, we move to a video-to-text setting and hypothesize that multimodal information from the full-length video and audio can directly facilitate abstractive dialogue summarization. We propose a parameter-efficient way for incorporating such information into a pre-trained textual summarizer and demonstrate improvements in the generated summaries.

Processing Multiword Expressions for Grammatical Error Correction

Friday, 3 February 2023, 12.00pm to 1.00pm
Speaker: Shiva Taslimipoor (University of Cambridge)
Venue: Computer Laboratory, Room FW09

Abstract:

Multiword expressions (MWEs) are combinations of two or more words with syntactic and semantic idiosyncratic behaviours. They are prevalent in any language and domain. Different categories of MWEs include idioms, nominal compounds, light verb constructions, verb particle constructions, and each pose particular challenges in processing. While they are known to be processed faster by native speakers, language learners find them difficult to understand and use. Like most machine translation (MT) systems, current Grammatical Error Correction (GEC) systems do not take them into consideration and are not good at correcting them.
In this talk, I give a brief presentation of a survey on different approaches to model MWEs in NLP. I summarise on how transformers advanced in dealing with them and what aspects are still lacking proper solutions. And finally, I illustrate our method for processing MWEs in a grammatical error correction system able to capture this type of errors better than baseline GEC systems.

Bio:

Shiva Taslimipoor is a postdoctoral research associate in the NLIP group at the University of Cambridge. She is a member of ALTA Institute where her research focus is on developing tools for automatic language learning and assessment. Before that she was a research associate at RGCL at the University of Wolverhampton where she completed her PhD in the area of Natural Language Processing on the topic of ‘Automatic Identification and Translation of Multiword Expressions’.

Next DLP: Cyber Security Talk - Well that's expensively weird - A deep dive into cloud incident response

Tuesday, 7 February 2023, 1.05pm to 1.55pm
Speaker: Tom Cope, CSO from NextDLP
Venue: FW11, William Gates Building

Tom Cope (https://tomcope.com/), CSO from NextDLP (https://www.nextdlp.com/) (AI powered Human-centered data loss protection solution) will be presenting a deep dive into investigating anonymous Google Cloud billing activity, which revealed great lessons in threat modelling, security design, threat hunting and incident response.

Please sign-up at the following link to attend: https://forms.gle/NVYrxHZ4gs3jJqki9

Some catering will be provided before the talk.

The unbreakable lightness of single neuron non-linearities in learning

Tuesday, 7 February 2023, 1.30pm to 2.30pm
Speaker: Yasser Roudi, Kavli Institute for Systems Neuroscience
Venue: CBL Seminar Room (in person), Engineering Department, 4th floor Baker building

A large body of work in the theory of neural networks (artificial or biological) has been performed on neural networks comprised of simple activation functions, prominently, binary units. Analysing such networks has led to some general conclusions. For instance, there is long held consensus that local biological learning mechanisms such as Hebbian learning are very inefficient compared to iterative non-local learning rules used in machine learning. In this talk, I will show that when it comes to memory operations such a conclusion is an artefact of analysing networks of binary neurons: when neurons with graded response, more reminiscent of the response of real neurons, are considered, memory storage in neural networks with Hebbian learning can be very efficient and close to the optimal performance. Turning to artificial neural networks, I will discuss how non-linearities influence the ability of Restricted Boltzmann Machines to express probability distributions over the visible nodes and how this affects learnability in these machines.

Refs:
Schönsberg, F., Roudi, Y., & Treves, A. (2021). Efficiency of local learning rules in threshold-linear associative networks. Physical Review Letters, 126(1), 018301.
Bulso, N., & Roudi, Y. (2021). Restricted boltzmann machines as models of interacting variables. Neural Computation, 33(10), 2646-2681.