Skip to content

msahamed/machine_learning_earthquakes-nuclear-explosion

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

31 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Separate Earthquake Signals from Nuclear Explosions

Basic Information:

In this project, I develop machine learning models to seperate earthquake signals from nuclear explosions. Both earthquakes and nuclear explosions generate seismic waves that can be detected thousands of kilometers away. However, it is not always an easy task to separate them from each other. Although seismologists have developed methods, it is sometimes very challenging. For example, a recent 3.4 magnitude quake from North Korea was interpreted as 'suspected explosion' but later identified as natural. More details can be found here : http://www.independent.ie/world-news/asia-pacific/small-north-korea-earthquake-likely-natural-not-caused-by-nuclear-test-say-experts-36161584.html

The primary objective is to develop machine learning models to separate natural earthquakes from nuclear explosions.

For inquiry:

For anything about the implementations, please feel free to write me an email :

Sabber Ahamed msahamed@memphis.edu Center For Earthquake Research and Information (CERI) The University of Memphis 3890 Central Ave Memphis, TN 38152, USA

Notebook files :

  • benchmark-final_features.ipynb: To create bench marks with different algorithoms
  • classify_earthquake_nn.ipynb : Create neural network model
  • classify_earthquakes.ipynb : This is the main file where support vector machine was used.
  • extract_features_seismograms.ipynb : Using this module I extract features from seismograms
  • seismogram_analysis.ipynb : This module is for analysising seismograms

    Codes and libraries

    This project requires Python 2.7 or 3. I have Used python 3.0. The following Python libraries are also required:

  • NumPy
  • Pandas
  • matplotlib
  • scikit-learn
  • Keras

    Datasets

    Datasets are not included to this project due to size. Please email me msahamed@memphis.edu if you need the datasets.

    Bug reports

    Bug reports, comments, and suggestions are always welcome. The best the channel is to create an issue on the Issue Tracker here at the repository : https://github.com/msahamed/classify-earthquakes-nuclear-explosion

    License

    This program is free software: you can redistribute it and modify it under the terms of the MIT.

  • About

    This repository contains codes associated with machine learning aiming to separate earthquakes signal from nuclear explosions.

    Resources

    Stars

    Watchers

    Forks

    Releases

    No releases published

    Packages

    No packages published