Module interpret 

Model Interpretability and Explainability.

This module provides tools for understanding model predictions through feature attribution and explanation methods.

Methods

• SHAP (SHapley Additive exPlanations): Computes feature importance using Shapley values from cooperative game theory.
• Permutation Importance: Measures feature importance by shuffling features and measuring prediction change.
• Feature Contributions: Decomposes predictions into per-feature contributions.

Example

use aprender::interpret::{KernelSHAP, Explainer};

// Create explainer with trained model
let explainer = KernelSHAP::new(model, background_data);

// Explain a prediction
let shap_values = explainer.explain(&sample);

// shap_values[i] = contribution of feature i to the prediction

References

• Lundberg, S. M., & Lee, S. I. (2017). A Unified Approach to Interpreting Model Predictions. NeurIPS.
• Ribeiro, M. T., et al. (2016). “Why Should I Trust You?”: Explaining the Predictions of Any Classifier (LIME). KDD.

Structs

CounterfactualExplainer

Counterfactual explanation generator.

CounterfactualResult

Result of counterfactual explanation search.

FeatureContributions

Feature contribution analysis.

IntegratedGradients

Integrated Gradients for neural network attribution.

LIME

LIME (Local Interpretable Model-agnostic Explanations).

LIMEExplanation

LIME explanation result.

PermutationImportance

Permutation feature importance.

SaliencyMap

Saliency Maps for neural network visualization.

ShapExplainer

SHAP (SHapley Additive exPlanations) values for feature attribution.

Traits

Explainer

Trait for model explainers.