python-scikit-learnHow to get Elasticnet feature importance
For ElasticNext model we can use trained coefficients to evaluate feature importance:
from sklearn import datasets, linear_model, model_selection
X, y = datasets.load_diabetes(return_X_y=True)
X_train, X_test, y_train, y_test = model_selection.train_test_split(X, y)
model = linear_model.ElasticNet()
model.fit(X_train, y_train)
fi = model.coef_ctrl + cfrom sklearn importimport module from scikit-learn | datasets.load_diabetesloads sample diabetes database |
model_selection.train_test_splitsplits given | .ElasticNet(creates ElasticNet training model |
.fit(train model with a given features and target variable dataset | .coef_returns list of coefficients of a trained model |
Usage example
from sklearn import datasets, linear_model, model_selection
X, y = datasets.load_diabetes(return_X_y=True)
X_train, X_test, y_train, y_test = model_selection.train_test_split(X, y)
model = linear_model.ElasticNet()
model.fit(X_train, y_train)
fi = model.coef_
print(fi)output
[ 0.15675432 0. 3.33520417 2.10046188 0.44089256 0.27115722
-2.23505284 2.37683059 3.24743807 1.85651586]
Related
More of Python Scikit Learn
- Using OneClassSVM for outliers detection example
- How to get best parameters from grid search CV
- Feature selection using KBest and chi2
- Grid search CV example
- How to inverse transform PCA reduced data
- Feature importance example
- PCA dimensionality reduction example
- How to get model homogeneity score
- How to resample with replacement
- Using Ridge regression model example
See more codes...