Skip to main content

Speakers & Abstracts

Invited Speakers

  • Ayesha Dunk, Skills Team, The Alan Turing Institute
  • Graham Cole, Newcastle University
  • Dr Mike Croucher, MathWorks
  • Dr Peter Laflin, Morrisons
  • Professor Vania Dimitrova, University of Leeds
  • James Morgan, ASDA supermarkets
  • Professor Petar Milin, University of Birmingham
  • Richard Plant, Data Study Groups, The Alan Turing Institute

Lightening Talks Speakers - Local DS&AI Education Interest Group, University of Leeds

  • Professor Paul Baxter, Professor of Biostatistics & Education, School of Medicine
  • Professor Eric Atwell, Professor of Artificial Intelligence for Language, School of Computing
  • Dr Matt Bawn, Lecturer in Bacterial Genomics, School of Biology
  • Dr Martin Callaghan, Curriculum Redefined Lecturer, School of Computing

Discussion Groups Facilitators (physical)

  • Dr Luisa Cutillo
  • Professor Eric Atwell
  • Professor Paul Baxter
  • Richard Plant

Discussion Groups Facilitators (remote)

  • Andrew Bailey
  • Mary Paterson
  • Dr Noorhan Abbas
  • Ayesha Dunk


Taming Pandas and Tidy Parsnips: Teaching foundational data skills as part of an innovation curriculum - Graham Cole, University of Newcastle

With over 200k vacant data roles across the UK, together with business sentiment reflecting their struggle to find suitably experienced candidates, we’ve our work cut out for us in the data education space! An holistic solution is no doubt required; refreshing the way we approach data and its underpinning concepts within the FE sector, ensuring graduates leave university with the skills and competencies required to compete, empowering current employees to upskill and educating organisational leaders to make robust and transparent decisions – driving innovation, productivity and growth. Data is a broad church, tying together the complementary, yet discrete, technical disciplines of statistics and computer science. Mastering these crafts takes skill, determination and, above all, practice and should be considered a significant achievement. Yet this specialist expertise alone is not enough to transform data in its rawest forms into actionable decisions that can bring about positive change.
Making changes, whether within an existing enterprise, through higher research or the creation of a start-up requires a third paradigm, a set of skills and competencies to explore, design, build and test value.
This session will provide an overview of the presenter’s current practice; an intensive, immersive and experiential programme embedding businesses within student groups to work on a current data-centric challenge. It will highlight how the data curriculum is drawn into the wider innovation narrative and discuss current thinking on the evolution of the programme.

Data Science skills: An employer's perspective - Peter Laflin, Morrisons Supermarkets

Morrisons have grown their Data Science team to over 30 people in less than 3 years.  In the talk, Peter will talk about his experience of building a high performing team with the right technical skills and the right culture to deliver impact in a large retail environment.

Machine Learning Challenges in Industry - Michael Croucher, Mathworks

Students often like to focus on models and ideally those models will include buzzwords such as Deep learning, CNNs, transformers and so on. They often focus on trying to beat state of the art by a fraction of a percent using standard benchmark datasets, all of their analysis is in a Jupyter notebook and they fixate over what ‘best’ programming language they should use. At MathWorks, we work closely with industrial partners who are adopting AI as part of their processes and products. Industries such as automotive, aerospace and defense, medical devices, semiconductors, and energy production. Their ML problems rarely intersect with those listed above. In this talk, I’ll discuss some of the common issues we have found across many industries and how MathWorks’ technology and people currently address them. My hope is to start a dialogue with interested academics in how we can work more closely together to do even better.

Learning about AI: Critical View through the Lenses of the Council of Europe's Values (Human Rights, Democracy, and Rule of Law) - Vania Dimitrova, University of Leeds

The Council of Europe (CoE) is an international organisation aimed at upholding human rights, democracy and the rule of law in Europe. It advocates for the enforcement of select international agreements reached by its 46 member states on various topics, linking to its core values: human rights, democracy, and rule of law. The CoE has formed a Committee on Artificial Intelligence which is conducting “work to elaborate an appropriate legal framework on the development, design, and application of artificial intelligence, based on the Council of Europe’s standards on human rights, democracy and the rule of law.” The Council of Europe’s ‘Steering Committee for Education Policy and Practice’ launched a new project as part of the 2020-21 ‘Education for Democracy’ programme. This project, entitled ‘Artificial Intelligence and Education,’ is exploring how learning with AI, learning about AI, and preparing for AI will contribute towards the Council of Europe’s core values on human rights, democracy and the rule of law. As a member of the CoE working group on AI and Education, I will outline the work of the group and will present our findings on education of AI.

Data and Humanities: love it or loathe it? - Petar Milin, University of Birmingham

Has Data Science inherited a bad reputation from her older sister, Statistics, for telling only "lies, and damn lies"? Is it trustworthy only for so-called Hard Science problems, while considered unsuitable if not deceptive in Humanities and Social Sciences? All sciences, equally, care little about what is obvious or certain; they like to loiter on slippery slopes, in dim places, and under that shade of likelihood. This might be the reason for Jaynes (2003) to shout there is one logic of science – Probability, the thing at the heart of both Statistics and Data Science. Yet, there is a cline in compliance and in the pursuit of such stance in daily scientific practice, from hard to social to humanities. Partly, this could be due to tradition or the lack of it, which means that a researcher in the humanities or social sciences can only double the trouble: uncertain about the phenomenon, inexperienced in new data sciencing. Another piece of the puzzle could also be in the nature of the data, particularly in the humanities, which is typically small and untamed, i.e., unique. Simplified, standardised, averaged descriptions of profoundly important dimensions of our life appear simply uncompassionate. Davies (2017) pointed out that "reducing social and economic issues to numerical aggregates and averages seems to violate some people’s sense of political decency". We can, however, circumvent the humanistic and social-scientific loathe of Data Science, not by (en)forcing or (over)selling the necessity of change, but by finding a role for these broad scientific domains within and for Data Science. Their scepticism can be transformative for the better. And Fisher's non-mechanistic approach to data modelling, with cycles of probing and inferring (as opposed to Person's and Neyman's automatisation or algorithmisation), could be something soothing for Humanities and Social Sciences mindset.

Data Study Groups: Collaborative Knowledge Transfer - Richard Plant , Data Study Groups, Alan Turing Institute

Data Study Groups are a Turing-hosted venue for researchers from a broad range of fields to apply their knowledge to new problems posed by industry, government, and civil society groups. In this session I will talk about how to join DSGs at Turing and I will describe the environment and the tools we provide to our participants.



Teaching our online MSc in Artificial Intelligence at the University of Leeds - Eric Atwell, University of Leeds

I have taught Artificial Intelligence and Data Science for many years on campus at Leeds University, and at our China campus in Chengdu. In 2021 we launched a fully-online MSc in Artificial Intelligence; our first cohort is now working on their final AI Project module. As Programme Leader, I have overseen the design of fully-online teaching and learning resources for 10 modules, including 2 which I developed myself: Data Mining and Text Analytics, and AI Project. I also collected useful feedback from our AI MSc students. Here are my 6 recommendations for good practice in teaching a fully-online Masters module in AI and/or DS, based on my experience and on student feedback: 1- carefully select a leading textbook in the field of the module accompanied by online learning resources eg pptx slides for each chapter, as a core for the module, to ensure peer-reviewed quality of syllabus and content. Eg see student feedback: “…I liked that the course was structured as a series of long lecture-style videos followed by textbook readings. Having a concrete textbook to follow along is a definite plus.” 2- go beyond the textbook and make the module more specific to Leeds by adding related content on Leeds AI research, eg our EU-funded research project EDUBOTS on chatbots in Higher Education; and our AI research on deep learning transformers for understanding the Quran. 3- structure content delivery so each unit includes a range of modes: lecture video(s), textbook chapter and other background recommended reading, practical activity, discussion forum, overview webinar. E.g. student feedback: “… the variety of resources allowed flexibility to learn in different methods (reading material, listening to lectures and discussion as well as through research and practical implementation).” 4- ensure recorded video lectures and “live” webinars are delivered by experienced subject experts, who can interest and engage students in the knowledge. E.g. student feedback: “… The professor was clearly passionate, well informed, candid and overall engaging. I thought the content and webinars were engaging … I enjoyed Professor Atwell's delivery of both the online lectures and the webinars … I would like to thank Prof Eric for his engagement during this module. … Appreciate the prof's sense of humour.” 5- use a variety of assessment types across the programme. E.g student feedback: “… A written project as opposed to coding was a welcome change.” 6- I did get some negative student feedback as well, mainly wanting NLP topics to be covered in more detail. Unfortunately there is insufficient time and scope to cover all text analytics topics in depth in one 15-credit module. I recommend adapting your AI MSc programme according to customer feedback; e.g. developing an additional module in Natural Language Processing, to cover missing research-led aspects. My recommendations for online MSc education in Artificial Intelligence are NOT about using smart ed-tech (except maybe chatbots!), but about re-use of good ideas tried and tested in on-campus AI teaching.

Education and Training at Leeds Institute for Data Analytics - Paul Baxter, University of Leeds

Developing Digital Skills in the Undergraduate Biology Curriculum - Matt Bawn, University of Leeds

Biology is at the heart of many of the challenges facing society, such as global disease, the rise of antimicrobial resistance and world food security. Data is central to developing effective solutions to these challenges and more and more biological data is being produced and analysed. Dealing with this increasing scale and complexity in biological and omics data has only been possible through synergy with computational and data science, creating a new frontier of discovery and interdisciplinary research. It is vital, therefore, that new biologists are equipped with the necessary skills to enable them to approach these grand challenges as they continue through their university education, enter academic research or work in industry. As a cohort biologists may have significantly different experience of data science, computation and mathematics and it is important that new curricula are designed around enabling equitable and useful learning outcomes. We propose the design of a guided learning journey through Data Science with three key stages: Informed users, generators and influencers of Data. Throughout this journey new computational and data science approaches (dry lab) will be aligned to existing wet lab biological assays and analyses. We hope that this approach will foster interdisciplinarity, transferrable skills and lead to an increased self-efficacy amongst learners.

Curriculum Redefined, Data Science Education and Day 1 student skills - Martin Callaghan, University of Leeds

The University of Leeds is engaging on a major multi-year project to overhaul the undergraduate curriculum and as Data Science researchers and educators we have a great opportunity to work with colleagues to demonstrate the importance and applicability of tools and technologies that our students can take with them into their future careers. In this talk, I'll discuss some of the ideas we have in the School of Computing, how this is impacting what and how we teach and how this is helping us to work with students to translate our research.