https://github.com/oaslananka/earthquakepredictionml

A Machine Learning Project for Predicting Earthquake Magnitudes and Occurrences example.
https://github.com/oaslananka/earthquakepredictionml

Last synced: 4 days ago
JSON representation

A Machine Learning Project for Predicting Earthquake Magnitudes and Occurrences example.

• Host: GitHub
• URL: https://github.com/oaslananka/earthquakepredictionml
• Owner: oaslananka
• License: mit
• Created: 2024-05-30T11:20:38.000Z (6 months ago)
• Default Branch: main
• Last Pushed: 2024-05-30T12:07:54.000Z (6 months ago)
• Last Synced: 2024-05-31T13:18:52.938Z (6 months ago)
• Language: Python
• Size: 148 KB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Earthquake Prediction using Machine Learning

This project aims to predict earthquake magnitudes and occurrences using machine learning models.

## Project Structure

```ini {"id":"01HZ4MWPV1WXE217FHBXMVKS0E"}

earthquake-prediction-ml/

├── data/

│   └── earthquake_data.csv

├── src/

│   └── main.py

├── README.md

├── LICENSE

└── requirements.txt

```

## Installation

1. Clone the repository:

```sh {"id":"01HZ4MWPV22SCECAA8RZ3TFA9H"}

git clone https://github.com/oaslananka/EarthquakePredictionML.git

cd EarthquakePredictionML

```

2. Install the dependencies:

```sh {"id":"01HZ4MWPV22SCECAA8S2CKJ758"}

pip install -r requirements.txt

```

## Usage

1. Run the main script:

```sh {"id":"01HZ4MWPV22SCECAA8S41S7ZCC"}

python src/main.py

```

## Data Source

The earthquake data is fetched from the USGS (United States Geological Survey) API. The data includes information about earthquake magnitudes, locations, depths, and times.

## Algorithms and Methods

This project uses several machine learning algorithms for earthquake prediction, including:

- Random Forest Regressor

- Gradient Boosting Regressor

- XGBoost Regressor

The models are trained using features engineered from the raw earthquake data, such as statistical measures, zero-crossings, peak counts, FFT values, and the Hilbert transform's amplitude envelope.

## Example Project

This project serves as an example of using machine learning techniques for predicting natural events. It demonstrates data fetching, preprocessing, feature engineering, model training, and prediction.

## License

This project is licensed under the MIT License. See the [LICENSE](LICENSE) file for details.