REVIEW 7 cited by
Quantum Neuron: an elementary building block for machine learning on quantum computers
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Even the most sophisticated artificial neural networks are built by aggregating substantially identical units called neurons. A neuron receives multiple signals, internally combines them, and applies a non-linear function to the resulting weighted sum. Several attempts to generalize neurons to the quantum regime have been proposed, but all proposals collided with the difficulty of implementing non-linear activation functions, which is essential for classical neurons, due to the linear nature of quantum mechanics. Here we propose a solution to this roadblock in the form of a small quantum circuit that naturally simulates neurons with threshold activation. Our quantum circuit defines a building block, the "quantum neuron", that can reproduce a variety of classical neural network constructions while maintaining the ability to process superpositions of inputs and preserve quantum coherence and entanglement. In the construction of feedforward networks of quantum neurons, we provide numerical evidence that the network not only can learn a function when trained with superposition of inputs and the corresponding output, but that this training suffices to learn the function on all individual inputs separately. When arranged to mimic Hopfield networks, quantum neural networks exhibit properties of associative memory. Patterns are encoded using the simple Hebbian rule for the weights and we demonstrate attractor dynamics from corrupted inputs. Finally, the fact that our quantum model closely captures (traditional) neural network dynamics implies that the vast body of literature and results on neural networks becomes directly relevant in the context of quantum machine learning.
Forward citations
Cited by 7 Pith papers
-
Canonical quantization of neurons
Canonical quantization turns a neuron into an activation observable of a parameterized Hamiltonian, with hybrid algorithms for training on quantum data and numerics showing advantage over classical Ising neurons.
-
Selective Feature Re-Encoded Quantum Convolutional Neural Network with Joint Optimization for Image Classification
Selective re-encoding of top PCA features inside a QCNN, plus joint optimization of a PCA-based and an autoencoder-based QCNN, improves binary image classification accuracy on MNIST and Fashion-MNIST over the paper's ...
-
Dynamic Synaptic Modulation of LMG Qubits populations in a Bio-Inspired Quantum Brain
A collective all-to-all qubit model with a synaptic-depression-style feedback loop is shown numerically to regulate its population around half-filling and to slow its collective oscillations as feedback strength grows.
-
Pseudo quantum advantages in perceptron storage capacity
A quantum perceptron with tunable-frequency oscillating activation achieves higher storage capacity than classical perceptrons, but the gain arises solely from the activation function form.
-
Hybrid Quantum Neural Networks for Efficient Protein-Ligand Binding Affinity Prediction
A hybrid quantum-classical network matches or slightly beats classical baselines on protein-ligand binding affinity prediction while using fewer parameters.
-
Perspectives on Utilization of Measurements in Quantum Algorithms
A survey that categorizes quantum measurement uses into static circuits, dynamic circuits, and challenge-solving techniques, and argues measurements deserve more attention in algorithm design.
-
Quantum Powered Credit Risk Assessment: A Novel Approach using hybrid Quantum-Classical Deep Neural Network for Row-Type Dependent Predictive Analysis
A hybrid quantum-classical deep network, trained separately on agricultural and personal loan categories, reaches 81-83% accuracy on an imbalanced bank dataset, but without classical baselines or uncertainty estimates...
Discussion (0). Continue with ORCID to comment.