CQT researchers and their collaborator present a quantum speed-up for machine learning
One of the ways that computers ‘think’ is by analysing relationships within large sets of data. CQT’s Jansen (Zhikuan) Zhao, Anupam Prakash and their collaborator have shown that quantum computers can do one such analysis faster than classical computers, for a wider array of data types than was previously expected.
The team’s proposed ‘quantum linear system algorithm’ is published in the 2 February issue of Physical Review Letters. In the future, it could help crunch numbers on problems as varied as commodities pricing, social networks and chemical structures.
“The previous quantum algorithm of this kind applied to a very specific type of problem. We need an upgrade if we want to achieve a quantum speed up for other data,” says Jansen, who is corresponding author on the work.
That’s exactly what the team is offering. The CQT researchers began collaborating with Leonard Wossnig when he visited the Centre. He was then a Master’s Student at ETH Zurich. Jansen is a PhD student, and Anupam is a research fellow. Jansen’s PhD is with the Singapore University of Technology and Design.
The first quantum linear system algorithm was proposed in 2009 by a different group of researchers. That algorithm kick-started research into quantum forms of machine learning, or artificial intelligence.
A linear system algorithm works on a large matrix of data. For example, a trader might be trying to predict the future price of goods. The matrix may capture historical data about price movements over time and data about features that could be influencing these prices, such as currency exchange rates. The algorithm calculates how strongly each feature is correlated with another by ‘inverting’ the matrix. This information can then be used to extrapolate into the future.
“There is a lot of computation involved in analysing the matrix. When it gets beyond say 10,000 by 10,000 entries, it becomes hard for classical computers,” explains Jansen. This is because the number of computational steps goes up rapidly with the number of elements in the matrix: every doubling of the matrix size increases the length of the calculation eight-fold.
The 2009 algorithm could cope better with bigger matrices, but only if the data in them is what’s known as ‘sparse’. In these cases, there are limited relationships among the elements, which is often not true of real-world data.
Jansen, Anupam and Leonard present a new algorithm that is faster than both the classical and the previous quantum versions, without restrictions on the kind of data it works for.
As a rough guide, for a 10,000 square matrix, the classical algorithm would take on the order of a trillion computational steps, the first quantum algorithm some 10,000s of steps and the new quantum algorithm just 100s of steps. The algorithm relies on a technique known as quantum singular value estimation.
There have been a few proof-of-principle demonstrations of the earlier quantum linear system algorithm on small-scale quantum computers. Jansen and his colleagues hope to work with an experimental group to run a proof-of-principle demonstration of their algorithm, too. They also want to do a full analysis of the effort required to implement the algorithm, checking what overhead costs there may be.
To show a real quantum advantage over the classical algorithms will need bigger quantum computers. Jansen estimates that “We’re maybe looking at three to five years in the future when we can actually use the hardware built by the experimentalists to do meaningful quantum computation with application in artificial intelligence.”
Learn more: Quantum algorithm could help AI think faster
The Latest on: Machine learning
Descartes Labs unleashes machine learning on space data
on August 15, 2018 at 12:28 pm
Satellite imagery across the visual spectrum is cascading down from the heavens. The challenge is to figure out how to process it, learn from it—and monetize it. Advances in distributed processing and ... […]
LinkedIn releases new data for finding and recruiting machine learning talent
on August 15, 2018 at 11:50 am
As artificial intelligence becomes more popular, skills like machine learning and deep learning are becoming more in demand. New data released from LinkedIn shows job recruiters are fighting to find m... […]
Oracle offers GraphPipe spec for machine learning data transmission
on August 15, 2018 at 10:16 am
Oracle has developed an open source specification for transmitting tensor data, which the company wants to become a standard for machine learning. Called GraphPipe, the specification provides a protoc... […]
Oracle open sources GraphPipe, a new standard for machine learning models
on August 15, 2018 at 9:00 am
Machine learning is expected to transform industries. However, its adoption in the enterprise has been slower than some might expect because it's difficult for organizations to deploy and manage machi... […]
Looking for machine learning experts? LinkedIn data shows how to find them
on August 15, 2018 at 6:54 am
More and more businesses are looking for employees with skills in the emerging field of machine learning. Recruiting those employees, however, can be difficult, given how in-demand they are. LinkedIn ... […]
Oracle open sources Graphpipe to standardize machine learning model deployment
on August 15, 2018 at 6:09 am
Oracle, a company not exactly known for having the best relationship with the open source community, is releasing a new open source tool today called Graphpipe, which is designed to simplify and stand... […]
New interactive machine learning tool makes car designs more aerodynamic
on August 15, 2018 at 6:04 am
The new software instantly shows stream lines as well as pressure on the surface (color-coded) of interactively deformable shapes. Credit: Nobuyuki Umetani When engineers or designers want to test the ... […]
Machine Learning Could Help Identify Author of an Anonymous Code
on August 15, 2018 at 4:22 am
Researchers have found that machine learning can be used to help identify pieces of codes, binaries, and exploits written by anonymous programmers, according to Wired. In other words, machine learning ... […]
Oracle open-sources Graphpipe to make it easier to deploy machine learning models
on August 15, 2018 at 2:00 am
Oracle today open-sourced Graphpipe, a tool created to make it easy to serve machine learning models in the cloud made by popular frameworks like TensorFlow, MXNet, Caffe2, and PyTorch. Graphpipe was ... […]
How Figure Eight Speeds Up Machine Learning With Video Object Tracking
on August 14, 2018 at 4:24 pm
Machine-learning systems need to be shown video information over and over, so they can remember it all. This new platform creates training data up to 100 times faster than human-only solutions. It gen... […]
via Google News and Bing News