Feature selection using RFE #1831

Onelinerhub · Dec 16, 2022 · 5c4a54e · 5c4a54e
1 parent 579c35b
commit 5c4a54e
Showing 1 changed file with 37 additions and 0 deletions.
diff --git a/python-scikit-learn/feature-selection-using-rfe.md b/python-scikit-learn/feature-selection-using-rfe.md
@@ -0,0  1,37 @@
 # Feature selection using RFE
 
 ```python
 from sklearn import datasets, svm, feature_selection
 
 X, y = datasets.make_friedman1(n_samples=50, n_features=10)
 
 selector = feature_selection.RFECV(svm.SVR(kernel="linear"), step=1, cv=5)
 selector = selector.fit(X, y)
 
 ranks = selector.ranking_
 ```
 
 - `from sklearn import` - import module from [lib:scikit-learn](https://onelinerhub.com/python-scikit-learn/how-to-install-scikit-learn-using-pip)
 - `.make_friedman1(` - creates Friedman #1 regression dataset
 - `.RFECV(` - creates RFE feature selection model
 - `svm.SVR(` - we use SVR model as estimator for RFE
 - `.fit(` - train model
 - `selector.ranking_` - returns features rankings
 
 group: feature-selection
 
 ## Example: 
 ```python
 from sklearn import datasets, svm, feature_selection
 
 X, y = datasets.make_friedman1(n_samples=50, n_features=10)
 
 selector = feature_selection.RFECV(svm.SVR(kernel="linear"), step=1, cv=5)
 selector = selector.fit(X, y)
 print(selector.ranking_)
 ```
 ```
 [1 1 3 1 1 1 1 2 1 1]
 
 ```