UniLVQ is an open-source Python library that provides a unified, extensible, and user-friendly implementation of Learning Vector Quantization (LVQ) algorithms for supervised learning. It supports both classification and regression tasks, and is designed to work seamlessly with the scikit-learn API.
Built on top of NumPy and PyTorch, UniLVQ combines rule-based and neural-inspired LVQ variants, making it suitable for both research and practical applications.
- ✅ Unified base API compatible with
scikit-learn - ✅ Traditional rule-based LVQ variants: LVQ1, LVQ2.1, LVQ3, Optimized LVQ1:
Lvq1Classifier,Lvq2Classifier,Lvq3Classifier,OptimizedLvq1Classifier
- ✅ Loss-based LVQ models: GLVQ, GRLVQ, LGMLVQ (PyTorch-based):
GlvqClassifier,GlvqRegressor,GrlvqClassifier,GrlvqRegressor,LgmlvqClassifier
- ✅ Support for both classification and regression
- ✅ Built-in support for early stopping, metric evaluation, data scaling
- ✅ Modular design for easy extension and customization
- ✅ CI-tested, documented, and easy to use
| Type | Algorithms | Module |
|---|---|---|
| Rule-based LVQ | LVQ1, LVQ2.1, LVQ3, Optimized LVQ1 (Classifiers) | classic_lvq.py |
| Generalized LVQ | GLVQ (Classifier, Regressor) | glvq.py |
| Generalized Relevance LVQ | GRLVQ (Classifier, Regressor) | grlvq.py |
| Local Generalized Matrix LVQ | LGMLVQ (Classifier) | lgmlvq.py |
Please include these citations if you plan to use this library:
@software{thieu20250515UniLVQ,
author = {Nguyen Van Thieu},
title = {UniLVQ: A Unified Learning Vector Quantization Framework for Supervised Learning Tasks},
month = June,
year = 2025,
doi = {10.6084/m9.figshare.28802435},
url = {https://github.com/thieu1995/UniLVQ}
}Install the latest version from PyPI:
pip install unilvqVerify installation:
$ python
>>> import unilvq
>>> unilvq.__version__For classification problem using LVQ1 classifier:
from unilvq import Lvq1Classifier
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
# Load data
X, y = load_iris(return_X_y=True)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3)
# Train LVQ1 model
model = Lvq1Classifier(n_prototypes_per_class=1, learning_rate=0.1, seed=42)
model.fit(X_train, y_train)
# Evaluate
y_pred = model.predict(X_test)
print("Accuracy:", accuracy_score(y_test, y_pred))As can be seen, you do it like any model from Scikit-Learn library such as SVC, RF, DT,... Please read the examples folder for more use cases.
Documentation is available at: 👉 https://unilvq.readthedocs.io
You can build the documentation locally:
cd docs
make htmlYou can run unit tests using:
pytest tests/We welcome contributions to UniLVQ! If you have suggestions, improvements, or bug fixes, feel free to fork
the repository, create a pull request, or open an issue.
This project is licensed under the GPLv3 License. See the LICENSE file for more details.
- 🔗 Official source code repository
- 📘 Official document
- 📦 Download releases
- 🐞 Issue tracker
- 📝 Notable changes log
- 💬 Of 7546 ficial discussion group
Developed by: Thieu @ 2025