Base abstractions and foundation classes.
tuiml.base package provides the object-oriented foundation for the entire library. It defines the core interfaces that concrete components implement, along with the decorators (@classifier, @regressor, @clusterer, @associator, @kernel, ...) that register them with the component registry.This package is organized into several key abstraction layers:
Classifier, Regressor)
Clusterer, Associator) models, plus their registration decorators and registry lookup helpers (get_algorithm, list_algorithms, search_algorithms).
Preprocessor, Transformer).
Metric base class and shared validation/aggregation
tuiml.evaluation.metrics.
BaseSplitter, the interface behind the train/test
tuiml.evaluation.splitting.
BaseTuner, ParameterGrid, ParameterDistribution).
DataGenerator and its task-specific subclasses).
Kernel base class and @kernel decorator for
CachedKernel.
NearestNeighborSearch and probability
ProbabilityEstimator primitives shared across algorithms.Base classes and registration decorators for machine learning algorithms....
Base class for probability estimators....
Base classes for feature engineering operations....
Base classes for synthetic data generators....
Base class for SVM Kernel functions....
Base classes and utility functions for evaluation metrics....
Base class for nearest neighbor search algorithms....
Base classes for preprocessing operations....
Base classes for data splitting....
Base classes for hyperparameter tuning.