Diagnostic Suite
Performance and Residual AnalysisIntroductionClassification MetricsRegression MetricsPerformance Evaluation in MoDeVaResidual AnalysisResidual Analysis in MoDeVaPerformance ComparisonSupervised Learning for Residual AnalysisExamples
Weakness DetectionIntroductionSlicingWeakness Detection in MoDeVaError Slicing for Weakness DetectionWeakness ComparisonAdvanced Slicing
Error Decomposition (AMIF)Error-Aware Clustering (Random Forest and LTC)Weak-Cluster Detection and Repair (FuseKernel detection + Mixture of Experts)Examples
Underfitting and OverfittingEmpirical Risk and Generalization GapSlicing Generalization GapOverfitting Slicing in MoDeVaOverfit ComparisonOverfitting and Model RobustnessRemediation Strategies for Model Weaknesses Identified by Gap AnalysisExamples
ReliabilityConformal PredictionReliability Analysis in MoDeVaIdentification of Reliability Issue and Impactful VariablesReliability Diagnostics in MoDeVaReliability ComparisonStrategies for Addressing Model WeaknessesSupervised Machine Learning: Random Forest ClusteringExamples
RobustnessInput Perturbation for Robustness TestRobustness Analysis in MoDeVaIdentification of Robustness Issue and Impactful VariablesRobustness ComparisonSupervised Machine Learning: Random Forest ClusteringStrategies for Addressing Model WeaknessesExamples
ResilienceResilience Analysis in MoDeVaMeasuring Distribution DriftStrategies for Addressing Model WeaknessesExamples
FairnessDefinitions of Group FairnessMetrics for Group FairnessFairness Metrics in MoDeVaFairness Evaluation in MoDeVaFairness ComparisonFairness MitigationExamples