Initial release of Adaptive Lasso - a scikit-learn compatible implementation of the Adaptive Lasso algorithm for feature selection and regression.
Features
- `AdaptiveLasso` estimator fully compatible with scikit-learn API
- Regularization path computation with `adalasso_path` function
Implementation Details
- Follows scikit-learn estimator interface (fit/predict)
Dependencies
- numpy (>=1.16.0)
- scikit-learn (>=0.24.0)
- matplotlib (>=3.0.0)
Installation
bash
pip install adalasso
References
This implementation is based on the methodology described in:
- Zou, H. (2006). The adaptive lasso and its oracle properties. Journal of the American Statistical Association, 101(476), 1418-1429.