DIKU Bits: Machine Learning for the Control of Spin Qubits



Oswin Krause, Tenure Track Assistant Professor in the Machine Learning (ML) section


Machine Learning for the Control of Spin Qubits


Spin qubits are one of the prime candidates for universal quantum computing. However, as spin qubits are the smallest qubit platform, they are most affected by noise and manufacturing imprecisions. For this reason, today single qubits are tuned by hand in a procedure that can take weeks. My goal in this lecture is to describe how machine-learning can be used to automate this task. for this, we will need to introduce what a qubit is and which operation we can apply to it and how those operations map onto the actual device.

Which courses do you teach?

I teach Numerical Optimisation (NO, MSc) and Probabilistic Machine Learning (PML,MSc)

Which technology/research/projects/startup are you excited to see the evolution of?

Spin-Qubits are now at the point where scaling up beyond the smallest lab devices becomes possible. I hope to see a quick scale up of the manufactured devices, which might hopefully give us Quantum Computing within the next 30 years!