Universal Approximation Theorem and error bounds for quantum neural networks and quantum reservoirs

Abstract

Universal approximation theorems are the foundations of classical neural networks, providing theoretical guarantees that the latter are able to approximate maps of interest. Recent results have shown that this can also be achieved in a quantum setting, whereby classical functions can be approximated by parameterised quantum circuits. We provide here precise error bounds for specific classes of functions and extend these results to the interesting new setup of randomised quantum circuits, mimicking classical reservoir neural networks. Our results show in particular that a quantum neural network with $\mathcal{O}({\epsilon }^{-2})$ weights and $\mathcal{O}(\lceil {\log }_2({\epsilon }^{-1})\rceil )$ qubits suffices to achieve accuracy $\epsilon >0$ when approximating functions with integrable Fourier transform.

References in corpus (20)

• Barren plateaus in quantum neural network training landscapes
• Recent Advances in Physical Reservoir Computing: A Review
• Quantum Circuit Learning
• The effect of data encoding on the expressive power of variational quantum machine learning models
• Power of data in quantum machine learning
• Data re-uploading for a universal quantum classifier
• Harnessing disordered quantum dynamics for machine learning
• Opportunities in Quantum Reservoir Computing and Extreme Learning Machines
• Temporal Information Processing on Noisy Quantum Computers
• Gaussian states of continuous-variable quantum systems provide universal and versatile reservoir computing
• Biology and medicine in the landscape of quantum advantages
• Universal Approximation Property of Quantum Machine Learning Models in Quantum-Enhanced Feature Spaces
• One qubit as a Universal Approximant
• Efficient decomposition of unitary matrices in quantum circuit compilers
• Expressivity of Quantum Neural Networks
• Learning Nonlinear Input-Output Maps with Dissipative Quantum Systems
• Towards Quantum Advantage in Financial Market Risk using Quantum Gradient Algorithms
• Quantum reservoir computing in finite dimensions
• Optimization of the Memory Reset Rate of a Quantum Echo-State Network for Time Sequential Tasks
• Randomness-enhanced expressivity of quantum neural networks

Cited by in corpus (1)

• High-expressibility Quantum Neural Networks using only classical resources