Predictive Analysis of Mushroom Toxicity Based Exclusively on Their Natural Habitat. حسب Enrico Bertozzi (22461709)

"…<br>The consistency of the results across different kernels demonstrates that the information contained in the habitat, by itself, leads to a very simple optimal decision rule (mostly the prediction of the most frequent class per habitat), which cannot be improved solely by model adjustments. …"

Supplementary Material 8 حسب Nishitha R Kumar (19750617)

"…In AMR studies, datasets often contain more susceptible isolates than resistant ones, leading to biased model performance. SMOTE overcomes this issue by generating synthetic samples of the minority class (resistant isolates) through interpolation rather than simple duplication, thereby improving model generalization.…"