Simultaneous multi-crop land suitability prediction from remote sensing data using semi-supervised learning

Model evaluation

The dataset is comprised of 2110 observations spanning 9 crops; more details on the dataset are presented in Table 3. The multi-crop model demonstrated an acceptable fit for all crops, except for corn and flaxseed (see Table 4). For corn, the number of unique districts in which it is grown is by far the smallest compared to the other crops (see column No. Unique Districts in Table 3), with corn being almost completely grown in just a small cluster in southern Manitoba, southern Ontario, and southern Quebec. Consequently, it is likely that there was not sufficient…