TY  - JOUR
PB  - SciPost Foundation
DO  - 10.21468/SciPostPhys.14.5.117
TI  - Quantum annealing for neural network optimization problems: A new approach via tensor network simulations
PY  - 2023/05/17
UR  - https://www.scipost.org/SciPostPhys.14.5.117
JF  - SciPost Physics
JA  - SciPost Phys.
VL  - 14
IS  - 5
SP  - 117
A1  - Lami, Guglielmo
AU  - Torta, Pietro
AU  - Santoro, Giuseppe E.
AU  - Collura, Mario
AB  - Here, we focus on the problem of minimizing complex classical cost functions associated with prototypical discrete neural networks, specifically the paradigmatic Hopfield model and binary perceptron. We show that the adiabatic time evolution of QA can be efficiently represented as a suitable Tensor Network. This representation allows for simple classical simulations, well-beyond small sizes amenable to exact diagonalization techniques. We show that the optimized state, expressed as a Matrix Product State (MPS), can be recast into a Quantum Circuit, whose depth scales only linearly with the system size and quadratically with the MPS bond dimension. This may represent a valuable starting point allowing for further circuit optimization on near-term quantum devices.
ER  -

@Article{10.21468/SciPostPhys.14.5.117,
	title={{Quantum annealing for neural network optimization problems: A new approach via tensor network simulations}},
	author={Guglielmo Lami and Pietro Torta and Giuseppe E. Santoro and Mario Collura},
	journal={SciPost Phys.},
	volume={14},
	pages={117},
	year={2023},
	publisher={SciPost},
	doi={10.21468/SciPostPhys.14.5.117},
	url={https://scipost.org/10.21468/SciPostPhys.14.5.117},
}