KEYWORD |
Neural networks for optimization: a quantum computing application
Reference persons BARTOLOMEO MONTRUCCHIO
Description This thesis settles in the context of a quantum computing application but does not require a deep knowledge of the underlying physics.
Among the currently available quantum architectures, the candidate will address neutral atoms quantum machines, such as Pasqal and QuEra technologies. These quantum architectures rely on Rydberg atoms that are positioned onto a 2D/3D register and interact, defining a unit-disk graph. To map computational problems of interest (e.g., combinatorial optimization problems) on these quantum machines, it is required to find the positions of the atoms and retrieve the interaction pattern of interest, still compliant with hardware constraints. Solving this embedding problem is equivalent to solving a constrained unit-disk graph problem (NP-hard) and requires the implementation of custom heuristics to scale with an increasing number of atoms.
In this regard, machine learning can come in handy, so the candidates will work on neural networks’ development to solve the embedding problem.
The pre-requirements to successfully address this thesis are knowledge about optimization, machine learning and sound Python or Julia programming skills.
The duration of the thesis work is expected to be around 6 months, adjustable based on the specific needs and skills.
Send CV to chiara.vercellino@linksfoundation.com and giacomo.vitali@linksfoundation.com specifying the thesis title.
Deadline 19/12/2023
PROPONI LA TUA CANDIDATURA