Deep Learning project to predict drought events in the Horn of Africa
# Predicting droughts at short-term scale
This repository contains a series of scripts to predict drought at short term scale using Machine Learning models.
## Table of Contents
- Installation
- Usage
- Contributing
- License
## Installation
Use the provided .toml file to install the python environment
## Usage
The script contains pipelines to download MSG SEVIRI images using the eumdac library and the EUMETSAT Data Tailor. Further, there is a pipeline to apply a Whittaker-filter from the modape library.
It contains an implementation of a ConvLSTM and a Graph Neural Network in the WaveNet architecture. The code is currently under development
## Contributing
In progress
## Acknowledgments
This project was supported by the Italian National PhD Program in Artificial Intelligence “AI & agrifood and environment” pillar. The research activity was conducted at the “CNR-ISAC” with the partnership and support of the University of Naples Federico II.
## License
This project is licensed under the MIT License.