VIP -- Variational Inversion Package with example implementations of Bayesian tomographic imaging
arXiv:2310.13325 · doi:10.26443/seismica.v3i1.1143
Abstract
Bayesian inference has become an important tool to solve inverse problems and to quantify uncertainties in their solutions. Variational inference is a method that provides probabilistic, Bayesian solutions efficiently by using optimization. In this study we present a Python Variational Inversion Package (VIP), to solve inverse problems using variational inference methods. The package includes automatic differential variational inference (ADVI), Stein variational gradient descent (SVGD) and stochastic SVGD (sSVGD), and provides implementations of 2D travel time tomography and 2D full waveform inversion including test examples and solutions. Users can solve their own problems by supplying an appropriate forward function and a gradient calculation code. In addition, the package provides a scalable implementation which can be deployed easily on a desktop machine or using modern high performance computational facilities. The examples demonstrate that VIP is an efficient, scalable, extensible and user-friendly package, and can be used to solve a wide range of low or high dimensional inverse problems in practice.
References in corpus (13)
- ADADELTA: An Adaptive Learning Rate Method
- Variational Inference: A Review for Statisticians
- MCMC using Hamiltonian dynamics
- Exchange Monte Carlo Method and Application to Spin Glass Simulations
- Automatic Differentiation Variational Inference
- HypoSVI: Hypocenter inversion with Stein variational inference and Physics Informed Neural Networks
- Seismic tomography using variational inference methods
- 3D Bayesian Variational Full Waveform Inversion
- Stein Variational Gradient Descent With Matrix-Valued Kernels
- HMCLab: a framework for solving diverse geophysical inverse problems using the Hamiltonian Monte Carlo method
- Faster Uncertainty Quantification for Inverse Problems with Conditional Normalizing Flows
- Wave-equation-based inversion with amortized variational Bayesian inference
- Amortized Normalizing Flows for Transcranial Ultrasound with Uncertainty Quantification