Monday, November 20
by Mariana Velho
The 7th International Conference on Quantum Techniques in Machine Learning starts at CERN
Today, Monday 20th of November, marked the launch of the 7th International Conference on Quantum Techniques in Machine Learning. This five-day event, being held at CERN, expects more than 300 attendees. Three pre-conference tutorials were held on Sunday the 19th of November.
Alberto Di Meglio, head of CERN IT Innovation and coordinator of CERN QTI Phase 1, during the opening session of the 7th International Conference on Quantum Techniques in Machine Learning at CERN.
This annual conference focuses on the interdisciplinary field of quantum technology and machine learning and aims to gather leading academic researchers and industry players to interact through a series of scientific talks focussed on the interplay between machine learning and quantum physics.
Co-organised by the CERN Quantum Technology Initiative (CERN QTI) – an initiative that investigates applications of quantum technologies for particle physics and beyond through comprehensive R&D and knowledge-sharing – the conference is sponsored by DESY, Intel, IBM Quantum, IONQ, PASQAL, Quantum AI, QUNOVA Computing, and the Università di Verona.
During the opening, Michele Grossi, CERN Quantum Senior Fellow, emphasised the journey from last year’s 6th edition of QTML, in Naples, where CERN was proposed as the next place to host the 7th edition of QTML, to finally being here today, “in such an important place not only for particle physicists but to every scientist, in the exact same room where the discovery of the Higgs boson was announced”.
Alberto Di Meglio, head of CERN IT Innovation and coordinator of CERN QTI Phase 1, traced the important path that has been made since the first CERN openlab workshop on Quantum Computing for HEP, in November 2018, the beginning of CERN QTI Phase 1, in January 2021, and the most recent announcement, in October 2023, of the Open Quantum Institute – an initiative born at GESDA, supported by UBS and to be hosted by CERN. The new CERN QTI phase 2 was also mentioned, highlighting its four centres of competence (“Hybrid Quantum Computing and Algorithms”; “CERN Quantum Technology Platforms”; “Quantum Networks and Communications”; and “Collaboration for Impact”), with the CERN IT department involved in three out of these four centres of competence.
Max Welling from the University of Amsterdam, giving a keynote talk on "A general message belief propagation framework for quantum computations"
Two highlights of the day were the keynote given by Max Welling, from the University of Amsterdam, and the invited talk by Nathan Killoran, from Xanadu Quantum Technologies Inc. After these, a group picture was taken during a coffee break, and more talks followed, this time focusing on areas such as “Quantum Learning and Quantum Advantage”, “Quantum Models and Data”, and “Generalisation”.
Nathan Killoran from Xanadu Quantum Technologies Inc, giving an invited talk on "Better than classical? The subtle art of benchmarking quantum models"
The first day finished on a high note with a private visit to the new CERN Science Gateway as a warm welcome to all the QTML 2023 participants.
Participants of the QTML 2023 conference at the “Discover CERN” exhibition in the newly launched CERN Science Gateway
Tomorrow, sessions will focus on areas such as “Architectures for Quantum Machine Learning”, “Symmetry and Geometric Quantum Machine Learning”, and “Trainability of Quantum”.
Visit the QTML 2023 and QTI websites for more details. More news from the conference will be posted, in the coming days, on QTML Daily Updates, QTI LinkedIn, QTI Twitter, and the computing-blog.
Tuesday, November 21
by Mariana Velho
Second day of the 7th International Conference on Quantum Techniques in Machine Learning at CERN
The 7th International Conference on Quantum Techniques in Machine Learning continued today, Tuesday 21st of November. Yesterday's sessions focused on areas such as “Quantum Learning and Quantum Advantage”, “Quantum Models and Data”, and “Generalisation”.
Today's sessions included the following topics: “Architectures for Quantum Machine Learning”, “Symmetry and Geometric Quantum Machine Learning”, and “Trainability of Quantum Architectures”. Technical talks on these topics were given by representatives of PASQAL, the University of Trieste, EPFL, IBM, and many other leading research institutes.
Marco Cerezo, from Los Alamos National Laboratory, giving an invited talk
A highlight of the day was an invited talk by Marco Cerezo, from Los Alamos National Laboratory, entitled “A unified theory of Barren plateaus for deep parametrized quantum circuits”. During his talk – and while warning us to brace ourselves because winter is coming! – Marco mentioned that Neural Networks are widely used today, but their development has seen different periods of stagnation (or winters!). Emphasizing that “although Variational Quantum Computing holds the tremendous promise to achieve a quantum advantage – a computational speedup – we need to guarantee that our architectures will work when scaled up and applied to large, realistic problems”.
The day closed with another highlight: the first of two poster sessions, where more than 60 posters were presented, showcasing the latest developments in various areas of quantum research. In total, more than 100 posters are being presented in two sessions; two important networking moments where attendees can showcase their work and discuss it with the other attendees, in a valuable exchange of ideas and knowledge.
Tomorrow's sessions focus on areas such as “Quantum Learning and Shadows”, “Machine Learning for Quantum Science”, “Experimental Implementations”, and “Architecture for Quantum Machine Learning”.
Visit the QTML 2023 and QTI websites for more details. More news from the conference will be posted, in the coming days, on QTML Daily Updates, QTI LinkedIn, QTI Twitter, and the computing-blog.
Wednesday, November 22
by Mariana Velho
Third day of the 7th International Conference on Quantum Techniques in Machine Learning at CERN
At CERN’s main auditorium, the 7th International Conference on Quantum Techniques in Machine Learning continued today, Wednesday 22nd of November. In yesterday's sessions, work was presented on areas such as “Architectures for Quantum Machine Learning”, “Symmetry and Geometric Quantum Machine Learning”, and “Trainability of Quantum Architectures”.
There were various highlights today, with the first one being an invited talk by Natalia Ares, from the University of Oxford, on “Machine learning for tackling quantum device variability”.
The second highlight was the industry session where companies such as ESA, Google, IBM, Intel, IONQ, NASA, and PASQAL showcased to the participants the latest work they are developing in the area.
Christa Zoufal, from IBM, presented the mission of IBM Quantum in “bringing useful quantum computing to the world but also making the quantum world safe” emphasizing on how to deal with errors in quantum systems. Masako Yamada, from IONQ, presented IONQ's recent and future applications in quantum machine learning. For example, in the optimization of Airbus cargo loading or on quantum machine learning for image recognition where “IONQ quantum computers are working to reliably distinguish real-world images like road signs for machine learning”.
Bertrand Le Saux, from ESA, presented how they are bringing the power of quantum computing into earth observation, “accelerating the future of earth observation via transformative/disruptive innovation, strengthening Europe’s world-leading competitiveness”. Jarrod McClean, from Google, highlighted the team priorities, mentioning that “NISQ (Noisy Intermediate Scale Quantum) is still a target of interest, but robust error correction appears to open many more doors.” and that “application motivation for error correction and understanding of how quantum machine learning plays a role in the distant future, remains a priority”.
Gian Giacomo Guerreschi, from Intel, presented the Intel Quantum Software Development Kit (SDK) and the Intel Quantum Neural Networks. And Davide Venturelli, from NASA, presented the NASA Quantum AI Lab and partner’s activities. Mentioning the important mandate for NASA Quantum AI Lab to “determine the potential for quantum computation to enable more ambitious and safer NASA missions in the future”.
Finishing the industry session, Panagiotis Barkoutsos, from PASQAL, presented PASCAL's progress on Quantum Machine Learning. Referring that “PASQAL’s mission is to make the world and industry quantum-ready and develop concrete quantum solutions for their most valuable problems. Being, at the same time, eager to research towards quantum advantage and the impact it can have in our socio-economic world”.
Besides the invited talk and the industry session, today's sessions focused on areas such as “Quantum Learning and Shadows”, “Machine Learning for Quantum Science”, “Experimental Implementations”, and “Architecture for Quantum Machine Learning”. Technical talks were given by representatives of the Centre for Quantum Technologies, the University of Pavia, Leiden University, EPFL, the University of Edinburgh, and many other leading research institutes.
The day closed with a very special dinner – and the last highlight of the day! –where participants had the chance to network over a traditional fondue Swiss dinner with some rock & blues music by “The Waffle Makers”!
Tomorrow's sessions encompass areas such as “Quantum Algorithms”, “Quantum Kernels”, and “Quantum Machine Learning for Physics”.
Visit the QTML 2023 and QTI websites for more details. More news from the conference will be posted, in the coming days, on QTML Daily Updates, QTI LinkedIn, QTI Twitter, and the computing-blog.
