/ML_full_wave_Inversion

Machine learning as seismic velocity-model building method for full-waveform inversion

Primary LanguageMATLABBSD 3-Clause "New" or "Revised" LicenseBSD-3-Clause

Machine learning as seismic prior velocity-model building method for full-waveform inversion: A case study from Colombia

Authors: Ursula Iturraran-Viveros, Andres M. Munoz-Garcia, Octavio Castillo-Reyes and Khemraj Shukla

Sample MATLAB code to perform 1-D inversion using Non-Linear Least Squares and the Kennett analytical solution for 1-D layered medium (used as forward modeling tool). Since the data shown in the paper is proprietary information we only give an example that is not related to this information but it is suitable to illustrate how to use this code.

Dependencies

  • A recent MATLAB release (version >= R2017b)

  • Optimization Toolbox (MATLAB)

  • Signal Processing Toolbox (MATLAB)

Running the code

The main code driver is invoked as follows:

  $ run Inversion1D_VpNN.m