Quantum activation functions for quantum neural networks

Abstract

The field of artificial neural networks is expected to strongly benefit from recent developments of quantum computers. In particular, quantum machine learning, a class of quantum algorithms which exploit qubits for creating trainable neural networks, will provide more power to solve problems such as pattern recognition, clustering and machine learning in general. The building block of feed-forward neural networks consists of one layer of neurons connected to an output neuron that is activated according to an arbitrary activation function. The corresponding learning algorithm goes under the name of Rosenblatt perceptron. Quantum perceptrons with specific activation functions are known, but a general method to realize arbitrary activation functions on a quantum computer is still lacking. Here we fill this gap with a quantum algorithm which is capable to approximate any analytic activation functions to any given order of its power series. Unlike previous proposals providing irreversible measurement-based and simplified activation functions, here we show how to approximate any analytic function to any required accuracy without the need to measure the states encoding the information. Thanks to the generality of this construction, any feed-forward neural network may acquire the universal approximation properties according to Hornik's theorem. Our results recast the science of artificial neural networks in the architecture of gate-model quantum computers.

Optimal Tuning of Quantum Generative Adversarial Networks for Multivariate Distribution Loading

“…In recent years, huge efforts were made to improve classical machine learning through quantum computers. The quantum version of most machine learning algorithms was experimented with, ranging from support vector machines (SVM) [15] to perceptrons [16][17][18], from feed-forward neural networks [17,19,20] to reservoir computing [21], and tensor networks [22] in gate model quantum computers, to quantum restricted Boltzmann machines [23,24] by adiabatic quantum computers, respectively. More specifically, in the domain of generative networks [25,26], encouraging results were obtained in comparison with the classical version in terms of the required network size [27].…”