Quantum HyperNetworks: Training Binary Neural Networks in Quantum Superposition
This work addresses the problem of efficiently training binary neural networks for deployment on energy- and memory-limited devices, offering a novel quantum-based approach that could impact machine learning applications, though it is incremental as it builds on existing quantum and binary network concepts.
The authors tackled the challenge of training binary neural networks, which involve combinatorial optimization problems, by introducing quantum hypernetworks that unify parameter, hyperparameter, and architecture search in a single quantum optimization loop, demonstrating through classical simulations that it effectively finds optimal configurations with high probability on classification tasks like a Gaussian dataset and scaled-down MNIST.
Binary neural networks, i.e., neural networks whose parameters and activations are constrained to only two possible values, offer a compelling avenue for the deployment of deep learning models on energy- and memory-limited devices. However, their training, architectural design, and hyperparameter tuning remain challenging as these involve multiple computationally expensive combinatorial optimization problems. Here we introduce quantum hypernetworks as a mechanism to train binary neural networks on quantum computers, which unify the search over parameters, hyperparameters, and architectures in a single optimization loop. Through classical simulations, we demonstrate that our approach effectively finds optimal parameters, hyperparameters and architectural choices with high probability on classification problems including a two-dimensional Gaussian dataset and a scaled-down version of the MNIST handwritten digits. We represent our quantum hypernetworks as variational quantum circuits, and find that an optimal circuit depth maximizes the probability of finding performant binary neural networks. Our unified approach provides an immense scope for other applications in the field of machine learning.