We aim to set the global standard for a truly interdisciplinary approach to contemporary data-driven research challenges. Established in 2015, the Data Science Institute (DSI) has over 300 members and has raised £50 million in research grants.
In 2025, the Data Science Institute (DSI) at Lancaster University proudly marks its 10th anniversary. Since its founding in 2015, the DSI has established itself as a leading hub for cutting-edge research, interdisciplinary collaboration, and real-world impact in data science and artificial intelligence. Over the past decade, our researchers and partners have tackled some of the most pressing challenges in society, science, and industry—advancing the foundations of data science, fostering ethical and trustworthy AI, driving innovation across sectors and training 100s of data science practitioners.
As we celebrate this milestone, we reflect on the achievements of our vibrant research community and the transformative projects that have shaped the field. Looking ahead, the DSI remains committed to pushing the boundaries of data science and AI research, strengthening global collaborations, and supporting the next generation of data scientists.
The Data Science Institute (DSI) invites nominations for three prizes celebrating research excellence and inclusive impact across data science and AI. We particularly welcome applications from individuals from historically underrepresented and underserved backgrounds, and we will make adjustments for career breaks.
Who can nominate: Self-nominations and nominations by others are equally welcome. For those in leadership roles, please encourage and support others to apply.
Award for each prize: Certificate + Gift voucher (one prize per category).
Inclusive Selection Process: Applications from individuals from underrepresented and underserved backgrounds are particularly encouraged, and special consideration will be given to applicants who have faced systemic barriers in academia. Additional time will be allowed for career breaks. A wide variety of reasons will be recognised. For parental leave, we allow 18 months per child for birthing parents, 6 months per child for non-birthing parents by default, but please state if a longer duration is required.
(please see in show more about these prizes, including submission details)
We would like your suggestions for speakers for Autumn 2025 - please get in touch if you would like to present or have a nomination to make!
Data Dialogues is an informal, discussion-driven event where members of the DSI and the broader university community share insights into their work, spark interdisciplinary conversations and explore potential collaborations. The focus is on interactive engagement rather than formal presentations—so no slides (or just a few, if needed)! Instead, the idea is to introduce your work in an accessible way, followed by an open discussion and Q&A with attendees.
Get fresh perspectives and think about new ways of approaching your own research, meet new people and explore potential research collaborations. Come be part of the DSI community!
Title: A Mixed-Integer Linear Programming Framework for the Adversarial Training of Neural Networks
Abstract: The training of neural networks (NNs) is a necessary task to improve their generalisation ability, measured by their performance on unseen inputs. However, even trained NNs can be vulnerable to adversarial inputs, which are minimally perturbed versions of standard inputs that are incorrectly labelled by the NN. The adversarial training of NNs can help to increase their robustness and guard against adversarial attacks. Recently, mixed-integer linear programming (MILP) models have been presented which are used to model the process of classification through trained NNs. One prominent application of such models is the ability to generate adversarial examples through providing constraints on the target input while minimising the level of perturbation. MILP models have also been presented for training NNs, showcasing comparable accuracy to traditional stochastic gradient descent approaches.
In this talk, we use recent advances in the field of MILP to present the adversarial training of NNs as an optimisation problem. We present a number of settings for the presented framework which allow for training against various settings of adversarially generated inputs, with the goal of increased robustness at minimal cost to performance. Experimental results on the MNIST data set of handwritten digits evaluate the performance of the proposed approach, and we discuss how the framework fits within the state-of-the-art.
Theme: Environmental Data Science
Description: Join us on December 2nd, 2025 (9:30–13:00) at the Sky Lounge, InfoLab, for a vibrant hybrid symposium designed to bring together Early Career Researchers (ECRs) from across disciplines who are interested in environmental data science. This event offers a platform to share research, spark collaborations, and shape future initiatives supported by CEEDS and DSI.
The programme includes short talks, interactive discussions, and a group activity aimed at exploring key research questions, challenges in cross-disciplinary work, and opportunities for collaboration and funding.
A networking lunch will close the event.
If you would like to attend, please register in advance. ECRs interested in giving a short presentation (5–10 minutes) are warmly encouraged to submit a title and brief abstract (max 100 words) when registering.
Come connect, collaborate, and help shape the future of interdisciplinary research at Lancaster!
Please complete the order form if you would like to present a short talk.
Send your abstract to Julia Carradus j.carradus1@lancaster.ac.uk
Please sign up on Eventbrite
Data Science at Lancaster was founded in 2015 on Lancaster’s historic research strengths in Computer Science, Statistics and Operational Research. The environment is further enriched by a broad community of data-driven researchers in a variety of other disciplines including the environmental sciences, health and medicine, sociology and the creative arts.
Foundations research sits at the interface of methods and application: with an aim to develop novel methodology inspired by the real-world challenge. These could be studies about the transportation of people, goods & services, energy consumption and the impact of changes to global weather patterns.
The Health theme has a wide scope. Current areas of strength include spatial and spatiotemporal methods in global public health, design and analysis of clinical trials, epidemic forecasting and demographic modelling, health informatics and genetics.
Data Science has brought new approaches to understanding long-standing social problems concerning energy use, climate change, crime, migration, the knowledge economy, ecologies of media, design and communication in everyday life, or the distribution of wealth in financialised economies.
The focus of the environment theme has been to seek methodological innovations that can transform our understanding and management of the natural environment. Data Science will help us understand how the environment has evolved to its current state and how it might change in the future.
.jpg)
The Data Engineering theme aims to explore how we can utilise digital technologies to accelerate and enhance our research processes across the University.
.jpg)
More News
