2C. Quantum neouromorphic approaches to data analysis and inference

14:30 - 15:40, Aula 11

Organizer: Mauro Paternostro

Chair: G. Massimo Palma

Retrieving Past Quantum Features with Deep Hybrid classical-quantum Reservoir Computing

Gian Luca Giorgi

Abstract: Classical and quantum machine learning offer distinct strengths for complex problem-solving. We introduce deep classical-quantum reservoir computing, a hybrid approach that merges quantum reservoir computing (QRC) with a classical echo state network (ESN). This deep architecture leverages the coherent processing of QRC for quantum inputs and the powerful memory of ESNs. Our design overcomes limitations of individual approaches, enabling nonlinear processing of temporal quantum data and improved memory capabilities. This paves the way for tackling challenging tasks in areas like quantum information processing and complex time-series analysis.

Click here to view the abstract.

Quantum parameter estimation with quantum extreme learning machines

Luca Innocenti, Salvatore Lorenzo, Ivan Palmisano, Francesco Albarelli, Alessandro Ferraro, Mauro Paternostro and Massimo G. Palma

Abstract: Quantum extreme learning machines (QELMs) offer efficient data processing with input quantum states. We analyse the potential of QELM to characterize properties for observables of quantum systems. We establish parallels between QELM training and measurement reconstruction, and propose quantum shadow tomography for performance analysis. This offers insights and methods for practical applications and possible future experiments.

Click here to view the abstract.

Photonic memristor-based reservoir computing

Iris Agresti

Abstract: We discuss the pespectives for a photonic implementation of a reservoir computing architecture based on a recently demonstrated photonic quantum memristor.

Click here to view the abstract.