Modeling Species Distribution

Mélisande discusses the challenges of incorporating range maps and satellite imagery in species distribution modeling. While the model can identify species based on environmental data, the lack of comprehensive checklists in regions like Kenya limits accuracy. By implementing a weighted loss strategy to prioritize locations with more observations, the team aimed to enhance the reliability of their data, though results still lagged behind those from the US.