Seven papers authored by Computer Science researchers from Warwick have been accepted for publication at the 41st International Conference on Machine Learning, one of the top three global venues for machine learning research, which will be held on 21-27 July 2024 in Vienna, Austria:

• Agent-Specific Effects: A Causal Effect Propagation Analysis in Multi-Agent MDPs, by Stelios Triantafyllou, Aleksa Sukovic, Debmalya Mandal, and Goran Radanovic
• Dynamic Facility Location in High Dimensional Euclidean Spaces, by Sayan Bhattacharya, Gramoz Goranci, Shaofeng Jiang, Yi Qian, and Yubo Zhang (Accepted as a spotlight, among the top 13 percent of all accepted papers)
• High-Dimensional Kernel Methods under Covariate Shift: Data-Dependent Implicit Regularization, by Yihang Chen, Fanghui Liu, Taiji Suzuki, and Volkan Cevher
• Revisiting character-level adversarial attacks, by Elias Abad Rocamora, Yongtao Wu, Fanghui Liu, Grigorios Chrysos, and Volkan Cevher
• Reward Model Learning vs. Direct Policy Optimization: A Comparative Analysis of Learning from Human Preferences, by Andi Nika, Debmalya Mandal, Parameswaran Kamalaruban, Georgios Tzannetos, Goran Radanovic, and Adish Singla
• To Each (Textual Sequence) Its Own: Improving Memorized-Data Unlearning in Large Language Models, by George-Octavian Bărbulescu and Peter Triantafillou
• Towards Neural Architecture Search through Hierarchical Generative Modeling, by Lichuan Xiang, Łukasz Dudziak, Mohamed Abdelfattah, Abhinav Mehrotra, Nicholas Lane, and Hongkai Wen

The University of Warwick will launch a new Computer Science programme with the Kharkiv National University of Radio Electronics (NURE) this September. This announcement coincides with the second anniversary of the University of Warwick's official twinning with NURE (29 Mar). The full story is available hereLink opens in a new window.

“Digitally Empowering Young People: The Podcast” is a ground-breaking podcast series hosted by Dr. Roxanne BibizadehLink opens in a new window.

In this inaugural series, we delve into the pressing issue of technology-assisted child sexual abuse material, focusing particularly on the misleading term “self-generated”, which problematically places the blame on the victim. Through this series, we aim to raise awareness and spark vital conversations among educators, parents/carers, law enforcement agencies and professionals working with young people.

Each episode features a distinguished expert voice, offering invaluable insights and perspectives on this critical issue. Contributors include esteemed organisations such as the Internet Watch Foundation, Marie Collins Foundation, National Policing Vulnerable Knowledge and Practice Programme, Parent Zone, Kent County Council, and Voice Box.

Our final episode is created especially for young people, providing them with essential information and resources to navigate the digital landscape safely and responsibly.

To listen to our podcast series, visit us on Spotify: https://open.spotify.com/show/3OANje22oUK5X641ACmxZOLink opens in a new window

For more information about this project and to stay updated on our latest initiatives, please visit our website: www.deyp.orgLink opens in a new window

We’re proud to announce that this project is funded by the ESRC IAA.

As part of a 2021/2022 MEng group project, Horia Druliac, Matthew Bardsley, Chris Riches, and Christian Dunn implemented a fully functional end-to-end (E2E) verifiable online voting system and conducted a successful trial among the residents of New Town in Kolkata, India during the 2022 Durga Puja festival celebration. This was the first time an E2E online voting system was built and tested in India. The feedback was overwhelmingly positive. Full details about the implementation, the trial and the voter feedback are written in a paper, published in the Journal of Information Security and Application. A free version of the paper is available on IACR e-print as a technical report. Also, see the earlier news item about this Durga Puja trial.

Professor Feng Hao, who supervised this group project, commented: “This is great teamwork. The four MEng students worked relentlessly for nearly a year, with good assistance from Luke Harrison and Professor Bimal Roy. The e-voting system was developed at an industry standard and worked flawlessly during the Durga Puja trial. Several government officials from India also helped us, providing invaluable support for the trial. We sincerely thank them in the acknowledgement section of the paper.”

Ahead of World Cancer Day on 4 February, scientists are revealing a cutting-edge artificial intelligence (AI) tool designed to help grade cancer, by analysing cell division.

In numerous cancer types, counting the number of cells undergoing division, known as mitotic figures, serves as a key indicator of cancer aggressiveness, or grade. This information helps inform treatment pathways, making it a crucial asessment tool. Traditional mitosis counting is both time-consuming and plagued by poor reliability. To address this, scientists have developed a new tool, MitPro, which uses AI to count and profile mitosis.

Histofy, a spin-out company from The University of Warwick that is leading developer of AI solutions for pathology, has engineered the tool to accurately profile mitosis throughout the entire tumour sample. This identifies the most suitable areas for further analysis.

On this recent article in the Quanta magazine, Alex Dixon, who wrote in Haskell the first solver for the problem, commented:

For the past 50 years, Vector Addition Systems—a simple but powerful computational model—have been a topic of great interest in theoretical CS. The reachability problem in that model asks whether we can get from some configuration to another.

The problem sounds relatively easy on a first glance, and an exponential lower bound held firm for over 40 years. Work by excellent theoreticians, including familiar names from Warwick DCS, finally closed the difficulty of the problem in 2021, concluding that it is very, very difficult indeed.