By IEEE Spectrum
June 15, 2017
A startup incubator connected to the University of Toronto anticipates the launch of quantum machine learning startups in the next few years.
Credit: Richard Kail/Science Photo Library
Researchers think the field of merging quantum computing with machine learning may soon be ripe for commercialization, with a startup incubator connected to the University of Toronto in Canada anticipating the launch of quantum machine learning startups in a few years’ time.
Creative Destruction Lab director Peter Wittek says the concept gained traction with the proliferation of the HHL algorithm, which solves massive linear algebraic problems involving many degrees of freedom with potentially greater speed than any conventional supercomputer can provide.
Wittek thinks particularly promising near-term quantum machine learning applications could be found in industries such as medicine, transportation, or finance.
The generation of purely random numbers is one expected advantage of such systems, and Microsoft’s Nathan Wiebe says they should be especially effective with qubits used as input instead of 0s and 1s.
Still, Wittek does not expect a quantum machine-learning system to dislodge traditional machine-learning applications based on graphics-processing units anytime soon.
From IEEE Spectrum
View Full Article
Abstracts Copyright © 2017 Information Inc., Bethesda, Maryland, USA
No entries found
from Communications of the ACM: Artificial Intelligence http://ift.tt/2t6ZR1I