I'd recommend we simplify this to "Generative adversarial", because it's more common to see "Generative adversarial network" rather than "Generative adversarial model", though both are used.
Probably not a huge issue in terms of the results of our analysis (AFAIK nobody has used GANs to study or measure biodiversity yet), but it's a small thing that stood out as potentially being too narrowly defined.
Currently there's an entry for "Generative adversarial model" in our search terms: https://github.com/th2018/ML_BIODIVERSITY_PAPER_HITS/blob/master/ml_biodiversity_hits.csv
I'd recommend we simplify this to "Generative adversarial", because it's more common to see "Generative adversarial network" rather than "Generative adversarial model", though both are used.
Probably not a huge issue in terms of the results of our analysis (AFAIK nobody has used GANs to study or measure biodiversity yet), but it's a small thing that stood out as potentially being too narrowly defined.