Training an Ising Machine with Equilibrium Propagation
This work addresses the challenge of integrating supervised training methods with Ising machine physics, potentially enabling these hardware platforms for broader AI applications, though it appears incremental in adapting existing algorithms.
The authors tackled the problem of training Ising machines for supervised learning by applying the Equilibrium Propagation algorithm, achieving results comparable to software implementations on the MNIST dataset and demonstrating support for convolutional networks with minimal spins per neuron.
Ising machines, which are hardware implementations of the Ising model of coupled spins, have been influential in the development of unsupervised learning algorithms at the origins of Artificial Intelligence (AI). However, their application to AI has been limited due to the complexities in matching supervised training methods with Ising machine physics, even though these methods are essential for achieving high accuracy. In this study, we demonstrate a novel approach to train Ising machines in a supervised way through the Equilibrium Propagation algorithm, achieving comparable results to software-based implementations. We employ the quantum annealing procedure of the D-Wave Ising machine to train a fully-connected neural network on the MNIST dataset. Furthermore, we demonstrate that the machine's connectivity supports convolution operations, enabling the training of a compact convolutional network with minimal spins per neuron. Our findings establish Ising machines as a promising trainable hardware platform for AI, with the potential to enhance machine learning applications.