This challange aims to exploit the cutting-edge hyperspectral imaging and machine learning to estimat soil parameters from hyperspectral images.
https://ai4eo.eu/portfolio/hyperview-challenge/
This repository contains the 2nd place winning solution. Hyperview Challange notebook goes through all the steps till the final prediction.
- Data splits folder contains the training/validation splits.
- Models weights folder contains the estimated weights for all models.
- Slected features folder contains the best selected features from RFE step.