Abstract
Accelerated regularization algorithms for ill-posed problems have received much attention from researchers of inverse problems since the 1980s. The current optimal theoretical results indicate that some regularization algorithms, e.g. the \(\nu \)-method and the Nesterov method, are such that under conventional source conditions the optimal convergence rates can be obtained with approximately the square root of the iterations of those needed for the benchmark (i.e. the Landweber iteration). In this paper, we propose a new class of regularization algorithms with parameter n, called the Acceleration Regularization of order n (AR\(^n\)). Theoretically, we prove that, for an arbitrary number \(n>-1\), AR\(^n\) can attach the optimal convergence rates with approximately the \(n+1\) root of the iterations needed for the benchmark method. Moreover, unlike the existing accelerated regularization algorithms, AR\(^n\)s have no saturation restriction. Some symplectic iterative regularizing algorithms are developed for the numerical realization of AR\(^n\). Finally, numerical experiments with integral equations and inverse problems in partial differential equations demonstrate that, at least for \(n\le 2\), the numerical behavior of AR\(^n\) matches our theoretical findings, also breaking the practical acceleration capability of all existing regularization algorithms.
Similar content being viewed by others
Notes
Inequality (55) \(\Leftrightarrow \sqrt{(1- b_k - e_k)^2 + 4 b_k}< 1 + b_k + e_k - \frac{2 e_k}{1 + b_k + e_k} \Leftrightarrow 0 < \left( \frac{2 e_k}{1 + b_k + e_k} \right) ^2\).
All the computations were performed on a dual core personal computer with 16.00 GB RAM with MatLab version R2021b.
Note that our theoretical results only hold in the regime \(n>-1\). Nevertheless, numerical experiments for \(n\le -1\) are also presented.
For ill-posed problems, it is not true that the smaller \(\tau \) the better result due to the semi-convergence phenomenon (consequently, the least square solution is not a regularization solution).
References
Nashed, M.: A new approach to classification and regularization of ill-posed operator equations. In: Inverse and Ill-posed Problems (Sankt Wolfgang, 1986), volume 4 of Notes Rep. Math. Sci. Engrg., pp. 53–75. Academic Press, Boston (1987)
Tikhonov, A., Goncharsky, A., Stepanov, V., Yagola, A.: Numerical Methods for the Solution of Ill-Posed Problems. Kluwer, Dordrecht (1995)
Tikhonov, A., Leonov, A., Yagola, A.: Nonlinear Ill-Posed Problems. Volumes I And II. Chapman and Hall, London (1998)
Engl, H., Hanke, M., Neubauer, A.: Regularization of Inverse Problems. Springer, New York (1996)
Kaltenbacher, B., Neubauer, A., Scherzer, O.: Iterative Regularization Methods for Nonlinear Ill-Posed Problems. Walter de Gruyter GmbH & Co. KG, Berlin (2008)
Tautenhahn, U.: On the asymptotical regularization of nonlinear ill-posed problems. Inverse Probl. 10, 1405–1418 (1994)
Vainikko, G., Veretennikov, A.: Iteration Procedures in Ill-Posed Problems (In Russian). Nauka, Moscow (1986)
Zhang, Y., Hofmann, B.: On the second-order asymptotical regularization of linear ill-posed inverse problems. Appl. Anal. 99, 1000–1025 (2020)
Kindermann, S.: Optimal-order convergence of Nesterov acceleration for linear ill-posed problems. Inverse Probl. 37, 065002 (2021)
Neubauer, A.: On Nesterov acceleration for Landweber iteration of linear ill-posed problems. J. Inverse Ill Posed Probl. 25, 381–390 (2017)
Zhang, Y., Hofmann, B.: On fractional asymptotical regularization of linear ill-posed problems in Hilbert spaces. Fract. Calculus Appl. Anal. 22, 699–721 (2019)
Boţ, R., Dong, G., Elbau, P., Scherzer, O.: Convergence rates of first- and higher-order dynamics for solving linear ill-posed problems. Found. Comput. Math. 22,1567–1629 (2022). https://doi.org/10.1007/s10208-021-09536-6
Gong, R., Hofmann, B., Zhang, Y.: A new class of accelerated regularization methods, with application to bioluminescence tomography. Inverse Probl. 36, 055013 (2020)
Schock, E.: Approximate solution of ill-posed equations: arbitrarily slow convergence vs. superconvergence. Constr. Methods Pract. Treat. Integral Equ. 73, 234–243 (1985)
Mathé, P., Pereverzev, S.V.: Geometry of linear ill-posed problems in variable Hilbert scales. Inverse Probl. 19(3), 789–803 (2003)
Albani, V., Elbau, P., de Hoop, M.V., Scherzer, O.: Optimal convergence rates results for linear inverse problems in Hilbert spaces. Numer. Funct. Anal. Optim. 37, 521–540 (2016)
Rieder, A.: Runge–Kutta integrators yield optimal regularization schemes. Inverse Probl. 21, 453–471 (2005)
Hairer, E., Wanner, G., Lubich, C.: Geometric Numerical Integration: Structure-Preserving Algorithms for Ordinary Differential Equations, 2nd edn. Springer, New York (2006)
Devore, R., Lorentz, G.: Constructive Approximation. Springer, New York (1993)
Zhang, Y., Hofmann, B.: Two new non-negativity preserving iterative regularization methods for ill-posed inverse problems. Inverse Probl. Imaging 15, 229–256 (2021)
Lions, J., Magenes, E.: Non-Homogeneous Boundary Value Problems and Applications. Volumes I. Springer, Berlin (1972)
Belgacem, B.F., Fekih, H.E.: On Cauchy’s problem: I. A variational Steklov–Poincaré theory. Inverse Probl. 21, 1915–1936 (2005)
Belgacem, B.F.: Why is the Cauchy problem severely ill-posed? Inverse Probl. 23, 823–836 (2007)
Alessandrini, G., Rondi, L., Rosset, E., Vessella, S.: The stability for the cauchy problem for elliptic equations. Inverse Probl. 25, 123004 (2009)
Hanke, M.: Conjugate Gradient Type Methods for Ill-Posed Problems. Wiley, New York (1995)
Acknowledgements
This work was supported by the Shenzhen National Science Foundation (no. 20200827173701001), National Natural Science Foundation of China (no. 12171036), Natural Science Foundation of Guangdong Province (no. 2019A1515110971) and Beijing Natural Science Foundation (no. Z210001).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The author declares that he has no conflict of interest.
Compliance with standards of research involving animals
This article does not contain any studies involving animals performed by any of the authors.
Compliance with standards of research involving humans as subjects
This article does not contain any studies involving human participants performed by any of the authors.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix A: Proof of Proposition 1
If \(\lambda <t^{-(n+1)}\), it is clear that \(\psi \in \mathcal {S}_{C}\) with \(C=1\). Now, suppose that \(\lambda \in [t^{-(n+1)},\Vert A\Vert ^2]\) (without loss of generality, we can assume that \(t> \Vert A\Vert ^{-\frac{2}{n+1}}\)). Let \(\psi \in \mathcal {S}_{C}\) and \(\varphi \) be covered by \(\psi \). Then, we have
which implies that \(\varphi \in \mathcal {S}_{C'}\) with \(C'=C/{\underline{c}}\).
Furthermore, \(\psi \in \mathcal {S}^{g}_{\zeta }\) and that assumption that \(\varphi \) is covered by \(\psi \) implies that
from which we can determine the relation \(\varphi \in \mathcal {S}^{g}_{\zeta }\).
If the function \(\lambda \mapsto \frac{\psi (\lambda )}{\varphi (\lambda )}\) is increasing, (25) is certainly satisfied with \(\underline{c}=1\). Moreover, when \(\varphi (\lambda )\in \mathcal {S}_{C}\) is concave, the function \(\xi (\lambda ):= \frac{\lambda }{\varphi (\lambda )}\) is monotonically increasing, as \(\xi '= (\varphi -\lambda \varphi ')/\varphi ^2>0\). Consequently, \(\varphi \) is covered by the function \(\lambda \).
Appendix B: Proof of Lemma 1
(i) The relation \(\varphi ^H_p\in \mathcal {S}_{C}\) follows from the following inequality:
(ii) First, the inequality \(\varphi ^L_{\mu }(\lambda _1\lambda _2) \le g(\lambda _1) \varphi ^L_{\mu }(\lambda _2)\) follows from
with \(g(\lambda _1)=1+\lambda _1\).
Since \(\varphi ^L_{\mu }(\cdot )\) is concave, according to Proposition 1 it is covered by function \(\varphi _1(\lambda )=\lambda \). Consequently, it is easy to show that \(\varphi ^L_{\mu }\in \mathcal {S}_{C}\) with \(C= e (n+1)\), and \(\varphi ^L_\mu \in \mathcal {S}^{g}_{\zeta }\) with \(g(\lambda )=1+\lambda \) and \(\zeta (\lambda )=e^{-\frac{2\lambda ^{2}}{n+1}}\).
Appendix C Proof of Theorem 6
We first investigate the convergence-rate results of the best worst-case error in a special situation.
Lemma 4
Suppose \(\omega (\lambda ) > 0\) for all \(\lambda > 0\). If we choose for every \(\delta > 0\) the parameter \(t_\delta > 0\) such that
then there exists a constant \(C_7 > 0\) such that
where \(C_g\) is the constant defined in (19).
Moreover, there exists a constant \(C_A > 0\) such that, for all \(\delta > 0\),
Proof
First, we note that the function
is continuous and strictly decreasing, and satisfies \(\lim _{t\rightarrow \infty } \xi _{\delta }(t) = 0\) and \(\lim _{t\rightarrow 0} \xi _{\delta }(t) = \infty \). Therefore, we find for every \(\delta > 0\) a unique value \(t_{\delta } = \xi ^{-1}_{\delta }(\delta )\). Consequently, \(t_\delta \) is the unique solution of equation (C1).
Let \({\tilde{y}}\in \bar{B}_\delta (y)\) be fixed. Using the inequality (19) together with the identity \(A g(t,A^*A)=g(t,AA^*)A\), it follows that
From this estimate, we obtain, with the triangular inequality and with the definition (C1) of \(t_\delta \),
which is the upper bound (C2).
For the lower bound (C3), we write, similarly,
Note that for every \(t > 0\) there exists a large-enough number \(T_\delta > t_\delta \) such that \([T^{-(n+1)}_\delta ,t^{-(n+1)}_\delta ]\) contains at least one eigenvalue of \(AA^*\) and the inequality \(1- e^{- \frac{T^{-(n+1)}_{\delta }}{n+1} t^{n+1}_\delta } \ge C_A \) holds with a fixed small positive number \(C_A\), dependent on only the distribution of eigenvalues \(\{ \sigma ^2_j \}^{\infty }_{j=1}\) of \(AA^*\). For instance, if , we set \(T_\delta = \kappa _A t_\delta \), and, consequently, we can choose \(C_A = 1 - e^{-1/[(n+1)\kappa ^{(n+1)}_A]}\). To that end, we let
where \(\{\sigma ^2_j; v_j\}_{j=1}^\infty \) is the eigen-system of the compact operator \(AA^*\). Through the choice of \(T_\delta \), an element \(y^\delta _1\in \mathcal {Y}\) exists such that
Let us first consider the case where \(z_\delta \ne 0\). Then, if we set \({\tilde{y}} = y + \delta \frac{z_\delta }{\Vert z_\delta \Vert }\), the equation (5) becomes
Thus, we may drop the last term as it is non-negative, which gives us the lower bound
Now, from
for all \(\lambda \in [T^{-(n+1)}_\delta ,t^{-(n+1)}_\delta ]\), we can estimate further that
Now, since the first term is decreasing in t, and the second term is increasing in t, we can estimate the expression for \(t > t_\delta \) from below using the second term at \(t = t_\delta \), and for \(t \le t_\delta \) using the first term at \(t = t_\delta \):
which yields the required inequality (C3).
If \(z_\delta \), as defined by (6), happens to vanish, the same argument works with the non-zero element \({\tilde{z}}_\delta \) by (7) since the last term in (5) is zero for \({\tilde{y}}=y+\delta \frac{{\tilde{z}}_\delta }{\Vert {\tilde{z}}_\delta \Vert }\). \(\square \)
From Lemma 4, we now obtain an equivalence relation between the noisy and the noise-free convergence rates.
Proof
Using \(\varphi \in \mathcal {S}^{g}_\zeta \), we have \(\phi (\gamma t)\le g(\gamma ^{-(n+1)}) \phi (t)\), which implies that
and so, by setting \({\tilde{g}}(\gamma ) = \gamma ^{-(n+1)/2} \sqrt{ g(\gamma ^{-(n+1)})}, \delta = {\tilde{\phi }}(t)\) and \({\tilde{\gamma }} = {\tilde{g}}(\gamma )\), we obtain, together with the monotonic decreasing of \({\tilde{\phi }}(\cdot )\),
Thus, we have
where \(h({\tilde{\gamma }}) = {\tilde{\gamma }}[{\tilde{g}}^{-1}({\tilde{\gamma }})]^{\frac{n+1}{2}}\).
The equivalence of (b) and (c) follows from Theorem 5. It is sufficient to show the equivalence of (a) and (b).
In the case where \(\Vert x(t) - x^{\dagger }\Vert = 0\) for all \(t \ge C_6\) for some \(C_6 > 0\), the inequality (44) is trivially fulfilled for some \({\tilde{c}} > 0\). Moreover, according to (4), by picking \(t=C_6\), we obtain
which implies the inequality (43) for some constant \(c > 0\), since
and we have, according to the definition of the function \(\psi \) and the decreasing property of \({\tilde{\phi }}(\cdot )\), \(\psi (\delta ) \ge a\delta \) for all \(\delta \in (0,\delta _{0})\) for some constants \(a > 0\) and \(\delta _{0} > 0\).
Thus, we may assume that \(\Vert x(t) - x^{\dagger }\Vert > 0\) for all \(t > 0\).
First, let (44) hold. According to Theorem 5, the inequality (45) holds. Consequently, for arbitrary \(\delta > 0\), we use the parameter \(t_{\delta }\) defined in (C1). Then, the inequality (45) implies that
Consequently,
and therefore, using the inequality (C2) obtained in Lemma 4, we find, with (8),
which is the estimate (43) with \(c = (1+C_g){\tilde{c}}h({\tilde{c}}^{-1})\).
Conversely, if (43) holds, we can use the inequality (C3) of Lemma 4 to obtain, from the condition (43),
Thus, with the definition of \(\psi \), we have
or, equivalently,
So, finally, we obtain, with the inequality \(\phi (\gamma t)\le g(\gamma ^{-(n+1)}) \phi (t)\),
and, since this holds for every \(\delta \), (45) holds with \({\tilde{c}} = \frac{c}{C_A}\sqrt{g\left( \frac{c^2}{C^2_A}\right) }\). Consequently, (44) holds. \(\square \)
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
Zhang, Y. On the acceleration of optimal regularization algorithms for linear ill-posed inverse problems. Calcolo 60, 6 (2023). https://doi.org/10.1007/s10092-022-00501-5
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s10092-022-00501-5