Skip to content

Hypocentral earthquake location using Stein-variational gradient descent and deep neural network travel-time formulations

License

Notifications You must be signed in to change notification settings

interseismic/HypoSVI

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

43 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

HypoSVI: Hypocenter inversion with Stein variational inference and Physics Informed Neural Networks



Introduction

HypoSVI is a software package that allows for earthquake location by leveraging the travel-time formulations from a continuous deep neural network EikoNet formulation.

This approach is outline in greater detail in the publication: Smith et al. (2021) - HypoSVI: Hypocenter inversion with Stein variational inference and Physics Informed Neural Networks - link to paper

Colab Jupyter Notebook

We have provided a Colab notebook to allow uses to go through the examples outlined in the paper provided above. The Colab notebook is separated into a series of sections which are all standalone executable scripts, but require the download and build of the software given in the 'Introduction' section of the notebook. As the software develops we will include additional sections.

Link to the Colab can be found at: link

Guide to Training and Setup

This package will be continued to be developed with the inclusion of a stand alone manual package. This is expected to be added within the next couple of months.

Installation

The HypoSVI software can be installed by using

  python setup.py install

Developers

Corresponding email - [email protected]

Jonathan Smith - California Institute of Technology
Kamyar Azizzadenesheli - California Institute of Technology
Zachary Ross - California Institute of Technology
Jack Muir - California Institute of Technology

About

Hypocentral earthquake location using Stein-variational gradient descent and deep neural network travel-time formulations

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%