Unsupervised Machine Learning on a Hybrid Quantum Computer

By J. Otterbach, R. Manenti, N. Alidoust, A. Bestwick, M. Block, B. Bloom, S. Caldwell, N. Didier, E. Fried, S. Hong and 20 more. 2017

305 citations recorded by the source corpus.

Machine learning techniques have led to broad adoption of a statistical model of computing. The statistical distributions natively available on quantum processors are a superset of those available classically. Harnessing this attribute has the potential to accelerate or otherwise improve machine learning relative to purely classical performance. A key challenge toward that goal is learning to hybridize classical computing resources and traditional learning techniques with the emerging capabilities of general purpose quantum processors. Here, we demonstrate such hybridization by training a 19-qubit gate model processor to solve a clustering problem, a foundational challenge in unsupervised learning. We use the quantum approximate optimization algorithm in conjunction with a gradient-free Bayesian optimization to train the quantum machine. This quantum/classical hybrid algorithm shows robustness to realistic noise, and we find evidence that classical optimization can be used to train around both coherent and incoherent imperfections.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/31fbba3638f0d930a678d85247045ea5a1a4ec88

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

A generative modeling approach for benchmarking and training shallow quantum circuits (2018, methodology): “QAOA is used in combinatorial optimization to ﬁnd approximate solutions of a classical Ising model [9] and it has been demonstrated on a 19-qubit device for the task of clustering synthetic data [10].”
QAOA for Max-Cut requires hundreds of qubits for quantum speed-up (2018, background): “In practice, one needs (tens of) thousands of repetitions of the same quantum circuit to have enough statistics to estimate the objective function [26, 27].”
Classification with Quantum Neural Networks on Near Term Processors (2018, background): “In the near future we will have gate model quantum computers with a sufficient number of qubits and sufficiently high gate fidelity to run circuits with enough depth to perform tasks that cannot be simulated on classical computers [19][20][21][22][23].”
Parameterized quantum circuits as machine learning models (2019, methodology): “Similarly, the quantum approximate optimization algorithm (QAOA) [8] has been used to ﬁnd approximate solutions of classical Ising models [9] and clustering problems formulated as MaxCut [10].”
Challenges and opportunities in quantum machine learning (2022, methodology): “Similar to classical ML, there exists several different QML paradigms: supervised learning (task-based) [26– 28], unsupervised learning (data-based) [29, 30] and reinforced learning (reward-based) [31, 32].”
Cloud Quantum Computing of an Atomic Nucleus. (2018, methodology): “In this Letter we present a quantum computation of the deuteron, the bound state of a proton and a neutron, and we use only publicly available software and cloud quantum hardware (IBM Q Experience and Rigetti 19Q [20]).”
Electric field spectroscopy of material defects in transmon qubits (2019, background): “Prototype quantum processors comprising a few tens of coupled qubits have already demonstrated quantum simulations of small molecules,(3) error correction,(4) and complex algorithms.(5) However, progress during the past decade was mostly achieved by improved circuit designs that reduced the interact”
Barren plateaus in quantum neural network training landscapes (2018, methodology): “These approaches have even been extended to both quantum machine learning and error correction [6, 7, 20–23].”
Hierarchical quantum classifiers (2018, background): “Originally proposed for quantum chemistry and combinatorial optimization, these approaches have been recently investigated for supervised [25] and unsupervised [26, 27] machine learning.”
Self-verifying variational quantum simulation of lattice models (2018, methodology): “Quantum variational techniques have been developed in quantum chemistry and in the context of classical optimization 15–19 .”
Quantum Approximate Optimization Algorithm: Performance, Mechanism, and Implementation on Near-Term Devices (2018, background): “As experiments begin to test solving the MaxCut problems with quantum machines [6, 8], limited quantum coherence time and graph connectivity will be among the biggest challenges.”
Quantum approximate optimization of non-planar graph problems on a planar superconducting processor (2020, methodology): “This work builds on prior experimental demonstrations of QAOA on superconducting qubits [22–25], ion traps [26], and photonics systems [27].”
From the Quantum Approximate Optimization Algorithm to a Quantum Alternating Operator Ansatz (2017, background): “Applications and extensions of QAOA beyond optimization include state preparation [28] and machine learning [29,30].”
Potential of quantum computing for drug discovery (2018, methodology): “Drawing on the connection to MAXCUT, a clustering algorithm has been implemented [139] on a superconducting quantum device produced by Rigetti Computing, Inc.”