On Monday 28th August 2023, Ali Raza, a PhD student of the School of Computing and iCSS, passed his PhD defence successfully with minor corrections.
Ali is co-supervised by Professor Ludovic Koehl and Dr Kim Phuc Tran of ENSAIT (École nationale supérieure des arts et industries textiles), University of Lille in France and the iCSS Director Professor Shujun Li towards a cotutelle degree between the University of Lille and the University of Kent. His study was funded by a 3-year joint PhD scholarship, in the form of a PhD project “Smart Healthcare System with Federated Learning” (SHSFL). The PhD scholarship was funded by the project I-SITE Université Lille Nord-Europe 2021 of France under the grant number I-COTKEN-20-001-TRAN-RAZA (funding amount €135,000), and by the University of Kent in the form of a full overseas fee waiver (equivalent funding amount over £60,000). The I-SITE project is part of the grant (REF LABEX/EQUIPEX), a French State fund managed by the National Research Agency (ANR) under the frame program “Investissements d’Avenir” and the reference number I-SITE ULNE/ANR-16-IDEX-0004 ULNE.
His thesis and the defence were examined by the following five academics from the UK and France:
- Professor Patrick Siarry, Université Paris-Est Créteil, France,
- Dr Ramla Saddem, Université de Reims Champagne-Ardenne, France,
- Professor Hongmei (Mary) He, University of Salford, UK,
- Dr Rehmat Ullah, Cardiff Metropolitan University, UK, and
- Dr Peng Liu, University of Kent, UK.
Ali’s defence was organised following an agreement between the University of Lille and the University of Kent so that both sides’ procedures were followed. It took place as a hybrid event with Dr Peng Liu and Professor Shujun Li attending remotely and all others attended in person at ENSAIT in France. Ali gave a public presentation at the beginning of the defence, which was followed by examiners’ questions and his answers. The examiners praised Ali’s work and presentation highly, and found that only minor corrections are required to finalise his PhD thesis.
Ali’s PhD thesis is titled “Secure and Privacy-preserving Federated Learning with Explainable Artificial Intelligence for Smart Healthcare System”. Some representative papers and preprints Ali produced during his PhD study can be found below.
- Ali Raza, Kim Phuc Tran, Ludovic Koehl and Shujun Li, “AnoFed: Adaptive anomaly detection for digital health using transformer-based federated learning and support vector data description,” Engineering Applications of Artificial Intelligence, 121:106051, 16 pages, Elsevier, 2023
- Ali Raza, Kim Phuc Tran, Ludovic Koehl and Shujun Li, “Proof of Swarm Based Ensemble Learning for Federated Learning Applications,” in Proceedings of the 38th ACM/SIGAPP Symposium on Applied Computing (SAC 2023), pp. 152-155, ACM, 2023, full edition as a preprint arXiv.2212.14050
- Ali Raza, Kim Phuc Tran, Ludovic Koehl and Shujun Li, “Designing ECG Monitoring Healthcare System with Federated Transfer Learning and Explainable AI,” Knowledge-Based Systems, 236:107763, 17 pages, Elsevier, 2022
- Ali Raza, Shujun Li, Kim-Phuc Tran and Ludovic Koehl, “Detection of Poisoning Attacks with Anomaly Detection in Federated Learning for Healthcare Applications: A Machine Learning Approach,” arXiv:2207.08486 [cs.LG], first submitted on 18 July 2022 (v1), revised on 9 May 2023 (v2)
- Ali Raza, Kim Phuc Tran, Ludovic Koehl, Shujun Li, Xianyi Zeng and Khaled Benzaidi, “Lightweight Transformer in Federated Setting for Human Activity Recognition,” arXiv:2110.00244 [cs.CV], first submitted on 1 October 2021 (v1), last revised on 4 November 2022 (v3)
Many thanks to Ali’s examiners for spending time reviewing his thesis and attending his defence, and warm congratulations to Ali for the well-deserved success!
In addition to the good news of passing his PhD viva, Ali has also secured a position from the Honda Research Institute Europe (HRI-EU) in Germany to work on security and privacy of AI systems for Honda’s future vision. He will start the next page of his career later this year.
The PhD student Ali said, “I am deeply grateful for the invaluable experience of working within two university environments and exploring diverse cultures. My sincere thanks extend to my supervisors, Shujun Li, Ludovic Koehl, and Kim Phuc Tran, for their consistent support, expert guidance, and encouragement throughout my entire PhD journey. Their mentorship has profoundly enriched my research and contributed to both my personal and professional growth. I extend my gratitude to them and the academic community for fostering an environment that prioritizes learning and development, which has played a pivotal role in shaping my academic trajectory.”
Ali’s French co-supervisor Professor Ludovic Koehl said, “I would like to thank all the members of the Jury and the public attending Ali’s defence. Ali, this is not the time to talk about the content of your work, but I would particularly like to thank you for the 3 years I have spent working with you. I would like to emphasize that it has been a real pleasure to work with you, both scientifically and on a personal level, and that I have enjoyed every moment of our exchanges. You have great qualities and you’ve made a huge contribution to our research team, never measuring the time you’ve spent. I would like to thank you once again and wish you all the best for your professional career.”
Ali’s another French co-supervisor Dr Kim Phuc Tran said, “As the PI of the SHSFL project to which Ali’s doctoral thesis belongs and also a co-supervisor, I have had the pleasure of working with Ali for the past 3 years. I thank Ali for the wonderful work he has done throughout his study. In addition to his research work, he has also been involved in guiding a master’s student for his thesis, assisting me in Machine Learning courses for doctoral students at the University of Lille. I wish Ali a successful career.”
Ali’s UK co-supervisor Professor Li said, “It has been a great pleasure to have worked with Ali and his other two French co-supervisors on the PhD project. This opportunity allowed us to explore applications of federated learning, an emerging and more privacy-preserving AI technology, to digital health. Ali is a very independent and hard-working student, and I am very glad to see that he managed to graduate within 3 years as planned. I look forward to continuing collaborating with him in his future new role within HRI-EU, with which the University of Kent is developing new research collaboration on AI-assisted security analysis for the automotive industry.”