The objective of this challenge is to create a machine learning model to classify crops in india using sentinel 2 data.
The challenge is hosted on Zindi.
This solution should rank me in TOP 4 although the chosen submission file put me on Top 5.
All notebook are run in Google colab
Run AgriField_DataPreparation.ipynb to produce train and test data (saved in data folder)
.Expected runtime: around 50 minutes.
PS : the runtime could be reduced to 20 mins if features are created in one function (iterating over tiles only one time)
Run AgriField_Modelling.ipynb Expected run time : around 6 minutes
I tried to explain all the features in the notebboks.
Vegetation indices could be found here.
| NAME | ZINDI ID | GITHUB ID |
|---|---|---|
| IHEB CHAABANE | @Reacher | @Iheb-ch |