Skip to content

Curriculum Guide

Bosaj edited this page Sep 21, 2026 · 1 revision

Academic Curriculum & Laboratory Guide πŸ“‚

This guide presents the complete educational structure, competency mapping, and lab sequence for ENIAD Explainable AI (XAI) Suite.


πŸ“‘ Curriculum Matrix

Laboratory / Directory Topic Key Techniques & Deliverables
Cours/ Theoretical XAI Foundations Interpretability taxonomy, algorithmic fairness, fidelity vs. interpretability
TP1/ Partial Dependence Plots (PDP) Marginal effect analysis, 1D and 2D interaction visualizers
TP2/ Accumulated Local Effects (ALE) Handling correlated features, unbiased localized attribution
TP3/ SHAP (Shapley Values) Cooperative game theory, TreeSHAP, KernelSHAP, force & summary plots
TP4/ LIME Local Surrogates Sparse linear explainers, image and tabular perturbation sampling
TP5/ Deep Learning Saliency Grad-CAM, gradient-weighted class activation mapping
ELHADJI Oussama/ Research Project Report Detailed experimental validation, comparative evaluation matrix

πŸ”¬ Key Methodologies Detailed

1. Partial Dependence Plots (PDP)

Partial dependence plots show the marginal effect one or two features have on the predicted outcome of a machine learning model. A partial dependence plot can show whether the relationship between the target and a feature is linear, monotonic or more complex.

2. Accumulated Local Effects (ALE)

ALE plots describe how features influence the prediction of a machine learning model on average. ALE plots are faster and unbiased alternatives to partial dependence plots (PDPs) when features are correlated.

3. SHAP (SHapley Additive exPlanations)

SHAP explains the prediction of an instance $x$ by computing the contribution of each feature to the prediction using coalitional game theory.

4. LIME (Local Interpretable Model-agnostic Explanations)

LIME explains individual predictions by approximating black-box models locally with an interpretable surrogate model (such as a sparse linear regression or decision tree).