I am planning to conduct a study on crop yield at parcel level in Ethiopia. I want to use artificial neural nets to estimate crop yield from satellite imagery.I want to know:
1. High resolution satellite imagery I should use.
2. The type of artificial neural nets structure good for the task.
3. The period good for the study (after summer?)
4. Is NDVI (Normalized Difference Vegetation Index) closely related to yield?
5. Are there alternative methods to field data collection for crop yield in training?