Sign In

Communications of the ACM

ACM TechNews

Machine Learning, Meet Quantum Computing


Structure of a perceptron.

Researchers at the University of Pavia in Italy have constructed a single-layer neural network, deployed it on a quantum computer, and induced it to perform simple image processing tasks.

Credit: Sagar Sharma

Researchers at the University of Pavia in Italy have constructed the world's first single-layer neural network—a perceptron—deployed on a quantum computer, and induced it to perform simple image processing tasks.

The researchers used IBM's Q-5 Tenerife superconducting quantum processor, which can process five quantum bits (qubits) and be programmed over the Web by whomever writes a quantum algorithm.

The team's algorithm takes a classical vector, like an image, as an input, combines it with a quantum weighting vector, and then generates a 0 or 1 output.

The researchers demonstrated that the Tenerife processor can handle input of N=2, and said the quantum perceptron was useful "as an elementary nonlinear classifier of simple patterns."

Planned future tasks for the quantum perceptron include encoding the equivalent of gray-scale images, and combining quantum perceptrons into multilayered networks.

From Technology Review
View Full Article

 

Abstracts Copyright © 2018 Information Inc., Bethesda, Maryland, USA


 

No entries found