Unsupervised Classification of Quantum Data

  title={Unsupervised Classification of Quantum Data},
  author={Gael Sent'is and Alex Monras and Ramon Mu{\~n}oz-Tapia and John Calsamiglia and Emilio Bagan},
  journal={Physical Review X},
We introduce the problem of unsupervised classification of quantum data, namely, of systems whose quantum states are unknown. We derive the optimal single-shot protocol for the binary case, where the states in a disordered input array are of two types. Our protocol is universal and able to automatically sort the input under minimal assumptions, yet partially preserves information contained in the states. We quantify analytically its performance for an arbitrary size and dimension of the data… 

Figures and Tables from this paper

Measurement disturbance tradeoffs in three-qubit unsupervised quantum classification
Surprisingly, a range of strategies are found in which a non-trivial first classification does not affect the success rate of the second classification, and this tradeoff is fully characterised analytically.
Beyond the Swap Test: Optimal Estimation of Quantum State Overlap.
The statistics of the optimal measurement are derived and the optimal mean square error is computed in the asymptotic pointwise and finite Bayesian estimation settings and two strategies relying on the estimation of one or both states are considered and show that, although they are suboptimal, they outperform the swap test.
Quantum Autoencoders to Denoise Quantum Data.
A novel quantum autoencoder is developed that successfully denoises Greenberger-Horne-Zeilinger, W, Dicke, and cluster states subject to spin-flip errors and random unitary noise.
Fidelity bounds for storage and retrieval of von Neumann measurements
The main goal is to estimate an asymptotic behavior of the maximum value of the average fidelity function F for a general N → 1 learning scheme, which reveals a particular class of learning schemes, which is called the pretty good learning scheme.
Constructing a qubit POVM from quantum data
This work considers the case in which the Bloch vectors of the states have the same z component, and finds a POVM that will discriminate between the two states by measuring the qubits.
Quantum self-supervised learning
The first steps to understanding whether quantum neural networks (QNNs) could meet the demand for more powerful architectures are taken and a numerical advantage for the learning of visual representations using small-scale QNN over equivalently structured classical networks is observed.
Quantum machine learning of graph-structured data
This work considers graph-structured quantum data in the form of pairs of input and output quantum states associated with the vertices of a graph and explains how to systematically exploit this additional graph structure to improve quantum learning algorithms.
Denoising quantum states with Quantum Autoencoders - Theory and Applications
The QAE shows particularly interesting results, as it enables to perform an almost perfect reconstruction of noisy states, but can also, more surprisingly, act as a generative model to create noise-free GHZ states.
Training deep quantum neural networks
A noise-robust architecture for a feedforward quantum neural network, with qudits as neurons and arbitrary unitary operations as perceptrons, whose training procedure is efficient in the number of layers.
Comparison protocol is optimal for quantum discrimination of single-qubit unitary gates—two candidates, one quantum sample per candidate
It is proved the optimality of the comparison protocol, implying that only one of the two samples is needed to achieve an optimal discrimination of a single-qubit unitary gate with two candidates.


Quantum learning without quantum memory
A quantum learning machine for binary classification of qubit states that does not require quantum memory is introduced and shown to perform with the minimum error rate allowed by quantum mechanics
Quantum learning of coherent states
It is shown that an optimal global measurement, performed jointly over the signal and the training set, provides higher successful identification rates than any learning strategy based on first estimating the unknown amplitude by means of Gaussian measurements on theTraining set, followed by an adaptive discrimination procedure on the signal.
Quantum speed-up for unsupervised learning
It is explained how it is possible to accelerate learning algorithms by quantizing some of their subroutines by giving quantized versions of clustering via minimum spanning tree, divisive clustering and k-medians that are faster than their classical analogues.
Identification of malfunctioning quantum devices
It is found that the optimal probability for detecting the position of rank-one and rank-two Pauli channels can be achieved by product state inputs and separable measurements for any size of network, whereas for rank-three and general amplitude damping channels optimal identification requires entanglement with N ancillas.
Optimal Universal Learning Machines for Quantum State Discrimination
We consider the problem of correctly classifying a given quantum two-level system (qubit) which is known to be in one of two equally probable quantum states. We assume that this task should be
Optimal measurements for symmetric quantum states with applications to optical communication
The minimum probability of error (MPE) measurement discriminates between a set of candidate quantum states with the minimum average error probability allowed by quantum mechanics. Conditions for a
Optimality of square-root measurements in quantum state discrimination
Distinguishing assigned quantum states with assigned probabilities via quantum measurements is a crucial problem for the transmission of classical information through quantum channels. Measurement
Quantum algorithms for nearest-neighbor methods for supervised and unsupervised learning
This work presents quantum algorithms for performing nearest-neighbor learning and k-means clustering and proves upper bounds on the number of queries to the input data required to compute such distances and find the nearest vector to a given test example.
Optimal quantum learning of a unitary transformation
The problem of the optimal inversion of an unknown unitary evolution is addressed, showing also in this case the optimality of the 'measure-and-rotate' strategies and applying the result to the optimal approximate realignment of reference frames for quantum communication.
Decision problems with quantum black boxes
It is examined how to distinguish between unitary operators, when the exact form of the possible operators is not known, and by suitable selection of an input state and generalized measurement of the output state, both unambiguous and minimum-error discrimination can be achieved.