| Rule | Logical statement defining conditions based on input data to predict a specific class. |
| Local Rule | Rule explaining the model's prediction for a specific data sample by focusing on the conditions relevant to that particular instance. |
| Global rule | Rule providing a general explanation of the model's behavior across the entire dataset. |
| Antecedent | A condition or statement in a rule that must be true for the rule to be activated. |
| Fidel/Fidelity | Measures how well a rule or a set of rules aligns with the model’s predictions. |
| Accuracy | Proportion of correctly predicted classes compared to the total number of predictions. |
| Covered samples | Set of data samples that satisfy the conditions of a specific rule. |
| Discriminant hyperplanes | Boundaries in the feature space that separate different classes based on the conditions defined by a model. |