RAPID EARTHQUAKE EARLY WARNING SYSTEM MACHINE LEARNING FRAMEWORK

This work implements four different deep learning model to predict:

1. The arrival of the S-wave in a four different locations.
2. The magnitude of the S-wave.
3. The epicenter.
4. The depth of the earthquake.

Based on the following independent variable:

1. Ten different detecting stations's locations.
2. Ten different detecting stations's P-wave arrival time.

REQUIREMENTS

1. Having a dedicated NVIDIA GPU + CUDA 8.0

INSTALLING

$ pip install --user pipenv
$ pipenv install --skip-lock

RUNNING IT

$ pipenv run python ./driver.py

Our models start converging at a number of epochs = 2000. Thus to Change the number of epochs:

$ EPOCHS=3000 pipenv run python ./driver.py

AUTHORS

Hero	email
Vicente Adolfo Bolea Sanchez	[email protected]
Olzhas Kaiyrakhmet	[email protected]>