from sklearn.linear_model import *
from finol.utils import load_config
distiller_dict = {
###############################
# Classical linear regressors #
###############################
"LinearRegression": LinearRegression, # Ordinary least squares Linear Regression
"Ridge": Ridge, # Linear least squares with l2 regularization
"RidgeCV": RidgeCV, # Ridge regression with built-in cross-validation
"SGDRegressor": SGDRegressor, # Linear model fitted by minimizing a regularized empirical loss with SGD
######################################
# Regressors with variable selection #
######################################
"ElasticNet": ElasticNet, # Linear regression with combined L1 and L2 priors as regularizer
"ElasticNetCV": ElasticNetCV, # Elastic Net model with iterative fitting along a regularization path
"Lars": Lars, # Least Angle Regression model a.k.a
"LarsCV": LarsCV, # Cross-validated Least Angle Regression model
"Lasso": Lasso, # Linear Model trained with L1 prior as regularizer (aka the Lasso)
"LassoCV": LassoCV, # Lasso linear model with iterative fitting along a regularization path
"LassoLars": LassoLars, # Lasso model fit with Least Angle Regression a.k.a
"LassoLarsCV": LassoLarsCV, # Cross-validated Lasso, using the LARS algorithm
"LassoLarsIC": LassoLarsIC, # Lasso model fit with Lars using BIC or AIC for model selection
"OrthogonalMatchingPursuit": OrthogonalMatchingPursuit, # Orthogonal Matching Pursuit model (OMP)
"OrthogonalMatchingPursuitCV": OrthogonalMatchingPursuitCV, # Cross-validated Orthogonal Matching Pursuit model (OMP)
#######################
# Bayesian regressors #
#######################
"ARDRegression": ARDRegression, # Bayesian ARD regression
"BayesianRidge": BayesianRidge, # Bayesian ridge regression
#############################
# Outlier-robust regressors #
#############################
"HuberRegressor": HuberRegressor, # L2-regularized linear regression model that is robust to outliers
"QuantileRegressor": QuantileRegressor, # Linear regression model that predicts conditional quantiles
"RANSACRegressor": RANSACRegressor, # RANSAC (RANdom SAmple Consensus) algorithm
"TheilSenRegressor": TheilSenRegressor, # Theil-Sen Estimator: robust multivariate regression model
##################################################
# Generalized linear models (GLM) for regression #
##################################################
"PoissonRegressor": PoissonRegressor, # Generalized Linear Model with a Poisson distribution
"TweedieRegressor": TweedieRegressor, # Generalized Linear Model with a Tweedie distribution
"GammaRegressor": GammaRegressor, # Generalized Linear Model with a Gamma distribution
#################
# Miscellaneous #
#################
"PassiveAggressiveRegressor": PassiveAggressiveRegressor, # Passive Aggressive Regressor
}
[docs]class DistillerSelector:
"""
Class to select distiller for model economic distillation.
"""
def __init__(self) -> None:
self.config = load_config()
[docs] def select_distiller(self):
"""
Select a distiller based on the configuration provided in the class.
"""
distiller_cls = distiller_dict.get(self.config["INTERPRETABLE_ANALYSIS_CONFIG"]["DISTILLER_NAME"], None)
if distiller_cls is None:
raise ValueError(f"Invalid distiller: {self.config['INTERPRETABLE_ANALYSIS_CONFIG']['DISTILLER_NAME']}. Supported distillers are: {distiller_dict}")
return distiller_cls()
