From quantum feature maps to quantum reservoir computing: an applicative perspective

We explore the interplay between two emerging paradigms: reservoir computing (RC) and quantum computing (QC). We observe how quantum systems featuring beyond-classical correlations and vast computational spaces can serve as non-trivial, experimentally viable reservoirs for typical tasks in machine learning. With a focus on neutral-atom quantum processing units, we describe and exemplify a novel quantum RC (QRC) workflow. We conclude by exploratively discussing the main challenges ahead, while arguing how QRC can offer a natural candidate to push forward RC applications. This article is part of the discussion meeting issue 'Bits, neurons and qubits for sustainable AI'.