by Tingting Wang¹, Yinju Bian¹, Yixiao Zhang ¹, Xiaolin Hou¹

¹Institute of Geophysics, China Earthquake Administration, Beijing 100081, China.

Corresponding author affiliation and e-mail:

Yinju Bian

Institute of Geophysics, China Earthquake Administration. No.5 South Minzu Unversity Road, Haidian District, Beijing 100081, China

[email protected]

This repository contains the source code to perform prediction and evaluation with example data.

• file folder: source

• Classify_svm.py
  Python script containing the SVM function to predict earthquake events classes.

• Classify_xgboost.py
  Python script containing the XGBoost function to predict earthquake events classes.

• file folder: data

• example36.csv
  This file contains 36 feature extraction training sets for different types of seismic events;

• example201.csv
  This file contains 201 waveform spectrum training sets for different types of seismic events;

• svmmodel_36.pkl

  SVM Pre-trained model to predict seismic event based on 36 features trainning set.

• svmmodel_201.pkl

  SVM Pre-trained model to predict seismic event based on 201 features trainning set.

• xgbmodel_36.pkl

  xgboost Pre-trained model to predict seismic event based on 36 features trainning set.

• xgbmodel_201.pkl

  xgboost Pre-trained model to predict seismic event based on 201 features trainning set.

The code has been tested using packages of:

• Python (version 3.7)
• numpy (1.19.2)
• pandas (1.0.1)
• scipy (1.6.1)
• scikit-learn (0.24.0)
• matplotlib (3.1.3)
• joblib (0.15.1)
• xgboost(1.3.1)

Running the code Classify_svm.py will perform the prediction and evaluation. to predict the results of data.

Running the code Classify_xgboost.py will perform the prediction and evaluation. to predict the results of data.

The following legal note is restricted solely to the content of the named files. It cannot overrule licenses from the Python standard distribution modules, which are imported and used therein.

BSD 3-clause license

Copyright (c) 2021 Tingting Wang, Yinju Bian, Yixiao Zhang , and Xiaolin Hou. All rights reserved.