An efficient Quasi-Newton method for nonlinear inverse problems via learned singular values

Abstract

Solving complex optimization problems in engineering and the physical sciences requires repetitive computation of multi-dimensional function derivatives. Commonly, this requires computationally-demanding numerical differentiation such as perturbation techniques, which ultimately limits the use for time-sensitive applications. In particular, in nonlinear inverse problems Gauss-Newton methods are used that require iterative updates to be computed from the Jacobian. Computationally more efficient alternatives are Quasi-Newton methods, where the repeated computation of the Jacobian is replaced by an approximate update. Here we present a highly efficient data-driven Quasi-Newton method applicable to nonlinear inverse problems. We achieve this, by using the singular value decomposition and learning a mapping from model outputs to the singular values to compute the updated Jacobian. This enables a speed-up expected of Quasi-Newton methods without accumulating roundoff errors, enabling time-critical applications and allowing for flexible incorporation of prior knowledge necessary to solve ill-posed problems. We present results for the highly non-linear inverse problem of electrical impedance tomography with experimental data.

References in corpus (8)

• Deep D-bar: Real time Electrical Impedance Tomography Imaging with Deep Neural Networks
• Deep Learning Diffuse Optical Tomography
• Classification of stroke using Neural Networks in Electrical Impedance Tomography
• A Flexible Krylov Solver for Shifted Systems with Application to Oscillatory Hydraulic Tomography
• Open 2D Electrical Impedance Tomography data archive
• Learning and correcting non-Gaussian model errors
• Big in Japan: Regularizing networks for solving inverse problems
• Learned SVD: solving inverse problems via hybrid autoencoding

Cited by in corpus (4)

• Structural engineering from an inverse problems perspective
• Impedance-optical Dual-modal Cell Culture Imaging with Learning-based Information Fusion
• Deep Inertia $L_p$ Half-Quadratic Splitting Unrolling Network for Sparse View CT Reconstruction
• Reconstructing the kinetic chemotaxis kernel using macroscopic data: well-posedness and ill-posedness