highfis 0.11.0

highFIS is a comprehensive Python package for training and evaluating high-dimensional TSK fuzzy systems, built on PyTorch and compatible with the scikit-learn API.

highFIS

highFIS is a PyTorch-based framework for high-dimensional Takagi–Sugeno–Kang (TSK) fuzzy systems. It brings differentiable fuzzy inference, numerical stability, and sklearn-compatible estimators to both classification and regression. The library also includes DGTSK dynamic-gating models for feature and rule selection in high-dimensional fuzzy systems.

🚀 Overview

• Differentiable TSK fuzzy systems built for high-dimensional data.
• Supports both concrete model classes and sklearn-style estimator wrappers.
• Includes adaptive and gated inference variants for feature selection and sparse rule extraction.
• Designed for numerical stability with log-space and inverse-log defuzzifiers.

📦 Installation

Install from PyPI:

pip install highfis

🧠 Quick Start

from highfis import HTSKClassifierEstimator

clf = HTSKClassifierEstimator(
    n_mfs=4,
    mf_init="kmeans",
    epochs=150,
    learning_rate=1e-3,
    random_state=42,
)
clf.fit(X_train, y_train)
print(f"Test accuracy: {clf.score(X_test, y_test):.4f}")

highFIS integrates with sklearn.pipeline.Pipeline, GridSearchCV, and cross_val_score.

🧩 Model families

highFIS provides a full family of TSK models, each tuned for a specific high-dimensional inference strategy.

• TSK — vanilla TSK with product antecedent aggregation and sum-based normalization.
• HTSK — high-dimensional TSK with geometric mean aggregation and log-space normalization.
• LogTSK — log-domain inverse-log normalization for stable aggregation.
• DombiTSK — Dombi t-norm aggregation with a learnable shape parameter.
• ADMTSK — adaptive Dombi TSK with Composite Gaussian membership functions.
• AYATSK — Yager aggregation for flexible antecedent behavior.
• AdaTSK — adaptive softmin-style inference with dynamic rule weighting.
• FSRE-AdaTSK — adaptive model with gated feature selection and rule extraction.
• DG-TSK — double-gated training for feature selection and rule extraction.
• DG-ALETSK — adaptive Ln-Exp softmin with embedded feature and rule gates.
• HDFIS — high-dimensional inference with product DMF (HDFIS-prod) and minimum T-norm frozen antecedents (HDFIS-min).

Each family exposes classifier and regressor variants.

🔧 Core components

Models

• HTSKClassifier, HTSKRegressor
• TSKClassifier, TSKRegressor
• LogTSKClassifier, LogTSKRegressor
• DombiTSKClassifier, DombiTSKRegressor
• ADMTSKClassifier, ADMTSKRegressor
• AYATSKClassifier, AYATSKRegressor
• AdaTSKClassifier, AdaTSKRegressor
• FSREAdaTSKClassifier, FSREAdaTSKRegressor
• DGTSKClassifier, DGTSKRegressor
• DGALETSKClassifier, DGALETSKRegressor
• HDFISProdClassifier, HDFISProdRegressor
• HDFISMinClassifier, HDFISMinRegressor

Estimator wrappers

• HTSKClassifierEstimator, HTSKRegressorEstimator
• TSKClassifierEstimator, TSKRegressorEstimator
• LogTSKClassifierEstimator, LogTSKRegressorEstimator
• DombiTSKClassifierEstimator, DombiTSKRegressorEstimator
• ADMTSKClassifierEstimator, ADMTSKRegressorEstimator
• DGALETSKClassifierEstimator, DGALETSKRegressorEstimator
• DGTSKClassifierEstimator, DGTSKRegressorEstimator
• FSREAdaTSKClassifierEstimator, FSREAdaTSKRegressorEstimator
• AYATSKClassifierEstimator, AYATSKRegressorEstimator
• AdaTSKClassifierEstimator, AdaTSKRegressorEstimator
• HDFISProdClassifierEstimator, HDFISProdRegressorEstimator
• HDFISMinClassifierEstimator, HDFISMinRegressorEstimator

Building blocks

• Memberships: GaussianMF, TriangularMF, TrapezoidalMF, BellMF, SigmoidalMF, DiffSigmoidalMF, ProdSigmoidalMF, SShapedMF, LinSShapedMF, ZShapedMF, LinZShapedMF, PiMF, CompositeExponentialMF, GaussianPIMF
• Defuzzifiers: SoftmaxLogDefuzzifier, SumBasedDefuzzifier, LogSumDefuzzifier
• T-norms: prod, min, gmean, dombi, yager, yager_simple, ale_softmin_yager
• Rule base strategies: cartesian, coco, en, custom
• Persistence: checkpoint save/load and validation via highfis.persistence

🛠️ Training options

highFIS uses gradient-based optimization and supports:

• adaptive optimizers like Adam/W and standard SGD
• early stopping with validation
• uniform rule regularization for balanced rule activation
• custom T-norms, custom rule bases, and custom defuzzifiers

📚 Documentation

The published documentation is available at:

https://dcruzf.github.io/highFIS

Key reference pages:

• TSK Vanilla
• LogTSK
• AYATSK
• HTSK
• DombiTSK
• ADMTSK
• AdaTSK
• DGTSK
• DG-ALETSK
• FSRE-AdaTSK
• HDFIS

🧪 Testing & quality

Run the test suite with coverage:

hatch test -c --a

Format and lint the repository:

hatch fmt

Run static type checks:

hatch run typing

🤝 Contributing

Contributions are welcome! Please open issues or pull requests, and refer to our development guide in the documentation:

https://dcruzf.github.io/highFIS/latest/contributing/

📄 License

Distributed under the GPLv3.