autoqild.detectors

The detectors package provides a suite of classes designed for detecting information leakage in machine learning experiments.

The package includes various implementations of leakage detectors utilizing popular machine learning frameworks and mutual information estimation techniques.

Modules

autogluon_leakage_detector	A leakage detection class leveraging AutoGluon for hyperparameter optimization and model evaluation.
ild_base_class	Abstract base class that defines the structure and core methods for leakage detection algorithms.
mi_estimator_detector	Detects leakage by estimating mutual information using GMM or MINE estimators.
mlp_leakage_detector	Uses a Multi-Layer Perceptron (MLP) for detecting leakage using deep learning approaches.
random_forest_leakage_detector	A leakage detector that utilizes RandomForest models for robust and interpretable detection.
sklearn_leakage_detector	A versatile leakage detection class built on top of the scikit-learn framework, supporting multiple estimators.
tabpfn_leakage_detector	Uses the TabPFN model to detect information leakage, particularly in small tabular datasets.