Install Python 3.6, jupyter notebook, and run pip install -r requirements.txt
This repo is an exploratory effort to see how Machine Learning can be applied to the CITES Trade Database, with three main objectives and experiments, each represented in a jupyter notebook:
- Given a partial permit, can we figure out what category it belongs to?
- Given some trade data over time, can we predict future trade numbers, per country and per species?
- Given valid and invalid permits (raw data before validation steps), can we learn the sanity checks and validation processes to spot invalid permits automatically?