Glossario IA
Il dizionario completo dell'Intelligenza Artificiale
Intrinsically Interpretable Models
Algorithms whose structure is naturally understandable without additional transformations
Post-Hoc Techniques
Methods applied after training to explain complex opaque models
SHAP (SHapley Additive exPlanations)
Game theory-based approach to attribute the importance of each feature
LIME (Local Interpretable Model-agnostic Explanations)
Model-agnostic method explaining individual predictions through local approximation
Global Feature Importance
Techniques quantifying the overall impact of each variable on model performance
Partial Dependence Plots
Visualizations showing the marginal effect of a feature on predictions
Counterfactual Explanations
Generations of minimal examples showing the necessary changes to modify a prediction
Interpretability of Neural Networks
Specific techniques for visualizing and understanding the internal mechanisms of deep learning
Interpretable Association Rules
Extraction of comprehensible logical rules from complex models
Interpretability Evaluation Metrics
Quantitative frameworks for measuring the quality and fidelity of generated explanations
Interpretability Visualizations
Graphical tools transforming explanation metrics into intuitive representations
Causal Interpretability
Methods distinguishing correlation and causality in model explanations