Quantum Machine Learning
Quantum computing is nowadays one of the most promising technologies due to quantum mechanical principles such as superposition and entanglement. These principles are not available on classical computers and provide the basis for novel algorithms with potentially exponential runtime improvements. A very well-known example is the quantum algorithm for prime factorization presented by Peter Shor.
However, quantum computers available today do not have the size or quality required for such algorithms. A potentially promising approach is quantum machine learning, in which parts of classical machine learning algorithms are combined with components running on a quantum computer. These hybrid quantum-classical approaches make it equally possible to support small numbers of qubits as well as mitigation of noise effects.
The quantum machine learning team at SCC combines expertise from computer science, electrical engineering, theoretical physics, as well as experimental physics, and addresses not only the fundamental issues of learnability and trainability in hybrid quantum-classical workflows, but also focuses on implementability, applicability, and scalability of algorithms. We are currently investigating the influences of variational quantum kernels, data encoding, and various cost functions.
In addition, we are working in close collaboration with the Simulation and Data Life Cycle Lab Particle and Astroparticle Physics on the prototypical implementation and evaluation of the possibilities of using quantum algorithms for high energy physics.