To increse the score of the model we need to remove the features which are recursive. Removing recursive feature reduces the computational cost and increase the efficiency.
So this is the recipe on How we can do recursive feature elimination in Python.
from sklearn.datasets import make_regression
from sklearn.feature_selection import RFECV
from sklearn import linear_model
We have only imported datasets to import the datasets, RFECV and liner_model.
We have imported inbuilt boston dataset and stored data in X and target in y. We have also used print statement to print rows of the dataset.
dataset = datasets.load_boston()
X = dataset.data
y = dataset.target
We have used linear Regression as a model and RFECV is used for recursive feature elimination we have used negative mean squared error as a scoring with cross validation as 4. We have fit and transform rfecv.
ols = linear_model.LinearRegression()
rfecv = RFECV(estimator=ols, step=1, scoring="neg_mean_squared_error", cv=4, verbose=0, n_jobs=4)
rfecv.fit(X, y)
rfecv.transform(X)
print(rfecv)
print(rfecv.n_features_)
So the output comes as
RFECV(cv=4,
estimator=LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None,
normalize=False),
min_features_to_select=1, n_jobs=4, scoring="neg_mean_squared_error",
step=1, verbose=0)
6