Stochastic data-driven Bouligand–Landweber method for solving non-smooth inverse problems
-
Harshit Bajpai
Abstract
In this study, we present and analyze a novel variant of the stochastic gradient descent method, referred as Stochastic data-driven Bouligand–Landweber iteration tailored for addressing the system of non-smooth ill-posed inverse problems. Our method incorporates the utilization of training data, using a bounded linear operator, which guides the iterative procedure. At each iteration step, the method randomly chooses one equation from the nonlinear system with data-driven term. When dealing with the precise or exact data, it has been established that mean square iteration error converges to zero. However, when confronted with the noisy data, we employ our approach in conjunction with a predefined stopping criterion, which we refer to as an a priori stopping rule. We provide a comprehensive theoretical foundation, establishing convergence and stability for this scheme within the realm of infinite-dimensional Hilbert spaces. These theoretical underpinnings are further bolstered by a numerical experiment on a system of linearly ill-posed problems and by discussing an example that fulfills the assumptions of the paper.
Funding statement: Harshit Bajpai would like to acknowledge Science and Engineering Research Board (SERB), India for providing a PhD fellowship through grant CRG/2022/005491. Ankik Kumar Giri would also like to thank SERB for supporting his research work as a part of the project “Fast and efficient iterative methods for solving nonlinear inverse problems” under the grant CRG/2022/005491.
References
[1] A. Aspri, S. Banert, O. Öktem and O. Scherzer, A data-driven iteratively regularized Landweber iteration, Numer. Funct. Anal. Optim. 41 (2020), no. 10, 1190–1227. 10.1080/01630563.2020.1740734Search in Google Scholar
[2] C. Christof, C. Meyer, S. Walther and C. Clason, Optimal control of a non-smooth semilinear elliptic equation, Math. Control Relat. Fields 8 (2018), no. 1, 247–276. 10.3934/mcrf.2018011Search in Google Scholar
[3] C. Clason and V. H. Nhu, Bouligand–Landweber iteration for a non-smooth ill-posed problem, Numer. Math. 142 (2019), no. 4, 789–832. 10.1007/s00211-019-01038-6Search in Google Scholar
[4] C. Clason and V. H. Nhu, Bouligand–Levenberg–Marquardt iteration for a non-smooth ill-posed inverse problem, Electron. Trans. Numer. Anal. 51 (2019), 274–314. 10.1553/etna_vol51s274Search in Google Scholar
[5] H. W. Engl, M. Hanke and A. Neubauer, Regularization of Inverse Problems, Math. Appl. 375, Kluwer Academic, Dordrecht, 2000. 10.1007/978-94-009-1740-8_3Search in Google Scholar
[6] Z. Fu, W. Wang, B. Han and Y. Chen, Two-point Landweber-type method with convex penalty terms for nonsmooth nonlinear inverse problems, IMA J. Numer. Anal. 43 (2023), no. 2, 1115–1148. 10.1093/imanum/drac003Search in Google Scholar
[7] A. Hanafy and C. I. Zanelli, Quantitative real-time pulsed Schlieren imaging of ultrasonic waves, IEEE 1991 Ultrasonics Symposium, IEEE Press, Piscataway (1991), 1223–1227. 10.1109/ULTSYM.1991.234310Search in Google Scholar
[8] T. Jahn and B. Jin, On the discrepancy principle for stochastic gradient descent, Inverse Problems 36 (2020), no. 9, Article ID 095009. 10.1088/1361-6420/abaa58Search in Google Scholar
[9] B. Jin and X. Lu, On the regularizing property of stochastic gradient descent, Inverse Problems 35 (2019), no. 1, Article ID 015004. 10.1088/1361-6420/aaea2aSearch in Google Scholar
[10] B. Jin, Z. Zhou and J. Zou, On the convergence of stochastic gradient descent for nonlinear ill-posed problems, SIAM J. Optim. 30 (2020), no. 2, 1421–1450. 10.1137/19M1271798Search in Google Scholar
[11] Q. Jin, X. Lu and L. Zhang, Stochastic mirror descent method for linear ill-posed problems in Banach spaces, Inverse Problems 39 (2023), no. 6, Article ID 065010. 10.1088/1361-6420/accd8eSearch in Google Scholar
[12] B. Kaltenbacher, A. Neubauer and O. Scherzer, Iterative Regularization Methods for Nonlinear Ill-Posed Problems, Radon Ser. Comput. Appl. Math. 6, Walter de Gruyter, Berlin, 2008. 10.1515/9783110208276Search in Google Scholar
[13] F. Kikuchi, K. Nakazato and T. Ushijima, Finite element approximation of a nonlinear eigenvalue problem related to MHD equilibria, Japan J. Appl. Math. 1 (1984), no. 2, 369–403. 10.1007/BF03167065Search in Google Scholar
[14] P. Kügler, A derivative-free Landweber iteration for parameter identification in certain elliptic PDEs, Inverse Problems 19 (2003), no. 6, 1407–1426. 10.1088/0266-5611/19/6/010Search in Google Scholar
[15] P. Kügler, Convergence rate analysis of a derivative free Landweber iteration for parameter identification in certain elliptic PDEs, Numer. Math. 101 (2005), no. 1, 165–184. 10.1007/s00211-005-0609-2Search in Google Scholar
[16] S. Lu and P. Mathé, Stochastic gradient descent for linear inverse problems in Hilbert spaces, Math. Comp. 91 (2022), no. 336, 1763–1788. 10.1090/mcom/3714Search in Google Scholar
[17] G. Mittal and A. K. Giri, Nonstationary iterated Tikhonov regularization: Convergence analysis via Hölder stability, Inverse Problems 38 (2022), no. 12, Article ID 125008. 10.1088/1361-6420/ac99fbSearch in Google Scholar
[18] G. Mittal and A. K. Giri, A modified steepest descent method for solving non-smooth inverse problems, J. Comput. Appl. Math. 424 (2023), Article ID 114997. 10.1016/j.cam.2022.114997Search in Google Scholar
[19] F. Natterer, The Mathematics of Computerized Tomography, Classics Appl. Math. 32, Society for Industrial and Applied Mathematics, Philadelphia, 2001. 10.1137/1.9780898719284Search in Google Scholar
[20] G. Ólafsson and E. T. Quinto, The Radon Transform, Inverse Problems, and Tomography, Proc. Sympos. Appl. Math. 63, American Mathematical Society, Providence, 2006. 10.1090/psapm/063Search in Google Scholar
[21] J. Outrata, M. Kočvara and J. Zowe, Nonsmooth Approach to Optimization Problems with Equilibrium Constraints, Nonconvex Optimization and its Applications 28, Kluwer Academic, Dordrecht, 1998. 10.1007/978-1-4757-2825-5Search in Google Scholar
[22] J. Rappaz, Approximation of a nondifferentiable nonlinear problem related to MHD equilibria, Numer. Math. 45 (1984), no. 1, 117–133. 10.1007/BF01379665Search in Google Scholar
[23] H. Robbins and S. Monro, A stochastic approximation method, Ann. Math. Statistics 22 (1951), 400–407. 10.1214/aoms/1177729586Search in Google Scholar
[24] O. Scherzer, Convergence criteria of iterative methods based on Landweber iteration for solving nonlinear problems, J. Math. Anal. Appl. 194 (1995), no. 3, 911–933. 10.1006/jmaa.1995.1335Search in Google Scholar
[25] O. Scherzer, A modified Landweber iteration for solving parameter estimation problems, Appl. Math. Optim. 38 (1998), no. 1, 45–68. 10.1007/s002459900081Search in Google Scholar
[26] S. Tong, W. Wang and C. Dong, A data-driven Kaczmarz iterative regularization method with non-smooth constraints for ill-posed problems, Appl. Numer. Math. 192 (2023), 152–178. 10.1016/j.apnum.2023.06.006Search in Google Scholar
© 2025 Walter de Gruyter GmbH, Berlin/Boston
Articles in the same Issue
- Frontmatter
- Local convergence of the error-reduction algorithm for real-valued objects
- A rotation total variation regularization for full waveform inversion
- Stochastic data-driven Bouligand–Landweber method for solving non-smooth inverse problems
- On determining the fractional exponent of the subdiffusion equation
- Tow-parameter quasi-boundary value method for a backward abstract time-degenerate fractional parabolic problem
- Determining both leading coefficient and source in a nonlocal elliptic equation
- Discrete dynamical systems: Inverse problems and related topics
- Stability estimates for an inverse problem of determining time-dependent coefficients in a system of parabolic equations
- Numerical solutions to inverse nodal problems for the Sturm–Liouville operator and their applications
- Revisiting linear machine learning through the perspective of inverse problems
Articles in the same Issue
- Frontmatter
- Local convergence of the error-reduction algorithm for real-valued objects
- A rotation total variation regularization for full waveform inversion
- Stochastic data-driven Bouligand–Landweber method for solving non-smooth inverse problems
- On determining the fractional exponent of the subdiffusion equation
- Tow-parameter quasi-boundary value method for a backward abstract time-degenerate fractional parabolic problem
- Determining both leading coefficient and source in a nonlocal elliptic equation
- Discrete dynamical systems: Inverse problems and related topics
- Stability estimates for an inverse problem of determining time-dependent coefficients in a system of parabolic equations
- Numerical solutions to inverse nodal problems for the Sturm–Liouville operator and their applications
- Revisiting linear machine learning through the perspective of inverse problems
