Abstract
We propose models of quantum perceptrons and quantum neural networks based on Clifford algebras. These models are capable to capture geometric features of classical and quantum data as well as producing data entanglement. Due to their representations in terms of Pauli matrices, the Clifford algebras seem to be a natural framework for multidimensional data analysis in a quantum setting. In this context, the implementation of activation functions, and unitary learning rules are discussed. In this scheme, we also provide an algebraic generalization of the quantum Fourier transform containing additional parameters that allow performing quantum machine learning based on variational algorithms. Furthermore, some interesting properties of the generalized quantum Fourier transform have been proved.
Similar content being viewed by others
Data availability statement
No data is used to support this article.
Notes
If a operator A is Hermitian, then \(\exp (iA)\) is unitary.
References
Altaisky, M.V.: Quantum neural network, Technical Report, Joint Institute for Nuclear Research, Russia (2001). arXiv: quant-ph/0107012v2
Alves, R., Lavor, C.: Clifford Algebra applied to Grover’s Algorithm. Adv. Appl. Clifford Alg. 20, 477–488 (2010)
Bayro-Corrochano, E.: Applications of Geometric Algebra Vol. I. Computer Vision, Graphics and Neurocomputing, Chap 13, pp. 471–478. Springer, Berlin (2019)
Bayro-Corrochano, E., Gamboa-Soliz, S., Altamirano-Escobedo, G., Lechuga-Gutierres, L., Lisarraga-Rodriguez, J.: Quaternionic spiking and quaternionic quantum neural networks: theory and applications. Int. J. Neural Syst. 31, 2 (2021)
Beer, K., Bondarenko, D., Farrelly, T., Osborne, J., Salzmann, R., Scheiermann, D., Wolf, R.: Training deep quantum neural networks. Nat. Commun. 11(1), 808 (2020)
Benedetti, M., Lloyd, E., Sack, S., Fiorentini, M.: Parametrized quantum circuits as machine learning models. Quant. Sci. Technol. 4(4), 043001 (2019)
Biamonte, J., Wittek, P., Pancotti, N., Rebentrost, P., Wiebe, N., Lloyd, S.: Quantum machine learning. Nature 549, 195–202 (2017)
Bottaci, L.: Artificial neural networks applied to outcome prediction for colorectal cancer patients in separate institutions. Lancet 350(9076), 469–72 (1997)
Buchholz, S.: A Theory of Neural Computation with Clifford Algebras. PhD Thesis, Kiel (2005)
Buchholz, S., Sommer, G.: On Clifford neurons and multi-layer perceptrons. Neural Netws. 21, 925–935 (2008)
Cai, X.-D., et al.: Entanglement-based machine learning on a quantum computer. Phys. Rev. Lett. 1144, 110504 (2015)
Childs, A.M., Maslov, D., Nam, Y., Ross, N.J., Su, Y.: Toward the first quantum simulation with quantum speedup. Proc. Natl. Acad. Sci. 115(38), 9456–9461 (2017)
Ciresan, D., Meier, U., Schmidhuber, J.: Multi-column deep neural networks for image classification. In: IEEE Conference on Computer vision and Pattern Recognition, pp. 3642–3649 (2012)
da Silva, A.J., Ludermir, T.B., Oliveira, W.R.: Quantum perceptron over a field and neural network architecture in quantum computer. Neural Netw. 76, 55–64 (2016)
De Bie, H.: Clifford algebras, Fourier transforms, and quantum mechanics. Math. Methods Appl. Sci. 35(18), 2198–2228 (2012)
Ganesan, N.: Application of neural networks in diagnosing cancer disease using demographic data. Int. Comput. Appl. 1(26), 81–97 (2010)
Harrow, A.W., Hassidim, A., Lloyd, S.: Quantum algorithm for linear systems of equations. Phys. Rev. Lett. 103, 150502 (2009)
Horn, R.A., Johnson, C.R.: Matrix Analysis. Cambridge University Press, Cambridge (1990)
Horn, R.A., Johnson, C.R.: Topics in Matrix Analysis. Cambridge University Press, Cambridge (1991)
Jozsa, R., Miyake, A.: Matchgates and classical simulation of quantum circuits. Proc. R. Soc. A 464, 3089–3106 (2008)
Kak, S.C.: Quantum neural computing. Adv. Imaging Electron Phys. 94, 259–313 (1995)
Kerenidis, I., Landman, J., Mathur, N.: Classical and Quantum Algorithms for orthogonal neural networks (2021). arXiv:2106.07198v1 [quant-ph]
Lloyd, S., Mohseni, M., Rebentrost, P.: Quantum principal component analysis. Nat. Phys. 10, 631–633 (2014)
Lounesto, P.: Clifford Algebras and Spinors, 2nd edn. Cambridge University Press, Cambridge (2001)
Nielsen, M., Chuang, I.: Quantum Computation and Quantum Information. Cambridge University Press, Cambridge (2000)
Rosenblatt, F.: The Perceptron-a perceiving and recognizing automaton, Report 85-460-1. Cornell Aeronautical Laboratory (1957)
Schuld, M.: Petruccione. Machine Learning with Quantum Computers. Springer, Switzerland (2021)
Shao, Chapeng: A quantum model of feed-forward neural networks with unitary learning algorithms. Quant. Inf. Process. 19, 102 (2020)
Song, Hai-Jing., Zhou, D.L.: Group theory on quantum Boltzmann machine. Phys. Lett. A 399, 127298 (2021)
Teguri, T., Isokawa, T., Matsui, N., Nishimura H., Kamiura, N.: Time series prediction by quaternionic qubit neural network. In: International Joint Conference on Neural Networks (IJCNN), pp. 1–6 (2020)
Trindade, M.A.S., Pinto, E., Vianna, J.D.M.: Adv. Appl. Clifford Alg. 26, 771–792 (2016)
Vaz, J., Jr., Rocha, R., Jr.: An Introduction to Clifford Algebras and Spinors. Oxford University Press, New York (2016)
Vlasov, A.Y.: Clifford algebras and universal quantum gates. Phys. Rev. A 63, 054302 (2001)
Walter, M.: Symmetry and Quantum Information. Lecture Notes, Spring (2018)
Wehner, S., Winter, A.: Higher entropic uncertainty relations for anti-commuting observables. J. Math. Phys. 49, 062105 (2008)
Zhang, G., Rong, H., Neri, F., Prez-Jimnez, M.J.: An optimization spiking neural system for approximately solving combinatorial optimizations problems. Int. J. Neural Syst. 24(05), 1440006 (2014)
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest or competing interests.
Additional information
Communicated by Uwe Kaehler.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix
A Basic Elements of Clifford Algebras
In this appendix, we review some basic concepts about Clifford algebras [24, 32].
Given a vector space V, the Clifford algebra can be defined as quotient \(Cl(V,Q)=\frac{T(V)}{I}\), where \(I_{Q}\) is a two sided ideal generated by elements
for all \(v \in V\); Q is the quadratic form and T(V) is the tensor algebra. Alternatively, let V be a space vector over \(\mathbb {R}\) equipped with a symmetric bilinear form g, A an associative algebra with unit \(1_{A}\), and \(\gamma \) a linear application \(\gamma :V\rightarrow A\). The pair \((A, \gamma )\) is a Clifford algebra for the quadratic space (V, g) if A is generated as an algebra by \(\{\gamma (v); v \in V\}\), \(\{a 1_{A}; a \in \mathbb {R}\}\) and satisfies
for all \(v,u \in V\). Let V be a vector space \(\mathbb {R}^{n}\) and g a symmetric bilinear form in \(\mathbb {R}^{n}\) of signature (p, q) with \(p+q=n\). We will denote by \(Cl(p,q) \equiv Cl_{p,q} \equiv Cl(\mathbb {R}^{p,q})\) the Clifford algebra associated with the quadratic space \(\mathbb {R}^{p,q}\). In addition, we denote \(Cl(2n,0)\equiv Cl(2n)\). The even subalgebra is defined by:
where \(\widehat{\ \cdot \ }\) denotes graded involution, which keeps the sign of the elements belonging to even subspaces. Note that the involution of a k-vector \(\Gamma _{[k]}\) is given by \(\widehat{\Gamma }_{[k]}=(-1)^{k}\Gamma _{[k]}\). The groups spin(p, q) and \(spin_{+}(p,q)\) are given by:
and
respectively, where \(N(a)=\vert a \vert ^{2}=\langle \widetilde{a} a \rangle _{0}\) is related to the norm of elements of Clifford algebra, and \(\widetilde{\ \cdot \ }\) represents the reversion operator defined by \(\widetilde{\Gamma }_{[k]}=(-1)^{k(k-1)/2}\Gamma _{[k]};\) \(\langle \ \ \rangle _{k}:Cl_{p,q}\rightarrow \bigwedge _{k}(\mathbb {R}_{p,q})\), with \(\bigwedge _{k}(\mathbb {R}_{p,q})\) denoting the exterior algebra of vector space \(\mathbb {R}_{p,q}\).
The Lie Algebra of \(spin_{+}(p,q)\) is the space of bivectors \(\bigwedge ^{2} \mathbb {R}^{p,q}\). Let \(B_{1}\) and \(B_{2}\) be two bivectors. Then the commutator
is a bivector.
B Swap Test
The swap test consists of the decomposition [34]
where \(Sym^{2}(\mathbb {C}^{d})\) and \(\Lambda ^{2} (\mathbb {C}^{d})\) are symmetric and antisymmetric spaces, respectively. Consider the system in the state \(\vert 0, \phi , \psi \rangle \). The Hadamard gate transforms this state in
Then, the controlled swap gate produces
After a second application of Hadamard gate, we obtain
If we perform a measurement on the first qubit, we would get
By repeating N times this procedure, we obtain
This equation allows the calculation of quantum fidelity in a quantum computer through measurements on the first qubit.
C A Simple Example of Decomposition of Quantum Gates
Unitary operators can be obtained through exponentials of anti-Hermitian operators. We will show how a simple circuit can be built from exponentials of elements of representations of Clifford algebras following the prescription obtained in the reference [25]. Thus we will build a circuit related to the unitary transformation
where
and
\(U(\theta _{1})\) can be expressed as a product of two-level unitary gates
where
and
Analogously, \(U(\theta _{2})\) can be expressed as a product of two-level unitary gates
where
and
so that
The circuit is illustrated in the figure below.
where
and
Notice that
which is generally an entangled state. We will build states with the following notation:
where \(\theta _{1}=\theta \) and \(\theta _{2}=x\) and
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Trindade, M.A.S., Lula-Rocha, V.N.A. & Floquet, S. Clifford Algebras, Quantum Neural Networks and Generalized Quantum Fourier Transform. Adv. Appl. Clifford Algebras 33, 38 (2023). https://doi.org/10.1007/s00006-023-01279-7
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s00006-023-01279-7