Figure: Optimisation landscape of a classification task in quantum machine learning. Source: DLR
The figure shows the error rate as a function of the model parameters. It is a visual representation of the geometry of the underlying quantum optimisation problem with respect to the projected model parameters. A principal axis transformation is used to visually represent the high dimensional model parameter space. As the error rate for some model parameters is extremely high, they are clipped. The blue 'optimisation trajectory' represents the evolution of the model parameters during a machine learning process. The red point shows the final value.
The development of modern quantum computers is just now reaching a stage where the first applications will also offer practical quantum advantages in comparison to classical programmes. To support DLR's quantum initiative, the ELEVATE project (Enhanced probLEm solVing with quAntum compuTErs) provides a network to support scientists from all DLR institutes in the feasibility analysis of quantum computing projects. In addition to the Institute of AI Safety and Security, quantum computing experts from the Institute of Software Technology, the Institute of Quantum Technology, Space Operations and Astronaut Training, the Institute of Engineering Thermodynamics and the Remote Sensing Technology Institute are involved in ELEVATE. From January 2022, a series of theoretical and practical training events will be in place, and in parallel, short test projects will be under way, for which partners from other DLR institutes can apply for expert time. The aim of the project is to enable DLR institutes to work with quantum computers and to establish a strong position for DLR in the field of quantum computing methods.
Supporting DLR institutes in the application of quantum computers
Quantum computing experts provide methodological support in four areas:
Quantum machine learning applications
Combinatorial optimisation
Quantum simulation
Compilation and hardware
This expertise is imparted in a series of events, including lectures, workshops and hackathons, so that theoretical knowledge can be immediately put into practice. These series of events are repeated several times so that all interested parties at DLR can participate.
In parallel to the training series on quantum computing, exploration studies are being carried out with domain experts from other institutes. The focus is on assessing the feasibility and potential quantum benefits for specific projects and formalising the question in terms of quantum algorithms. On this basis, implementation can be addressed in follow-up projects. The duration of the exploration studies is six to twelve months.
Contribution Institute for AI Safety and Security
In the ELEVATE project, the Institute for AI Safety and Security is working on topics related to quantum AI. In particular, it is investigating the coding of classical image and time series data on quantum computers, tensor network-based AI approaches, and applications in fluid dynamics. The findings from the exploratory projects are fed back to the community in workshop formats to generate further ideas and find collaborations for follow-up projects.
