- Notifications
You must be signed in to change notification settings - Fork1
Before training a model or feed a model, first priority is on data,not in model. The more data is preprocessed and engineered the more model will learn. Feature selectio one of the methods processing data before feeding the model. Various feature selection techniques is shown here.
rakibhhridoy/MachineLearning-FeatureSelection
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
Step By step
- importing libraries & functions
importpandasaspdimportnumpyasnpimportmatplotlib.pyplotaspltfromsklearn.linear_modelimportLassofromsklearn.model_selectionimportGridSearchCVfromsklearn.feature_selectionimportSelectKBestfromsklearn.feature_selectionimportf_classiffromsklearn.ensembleimportExtraTreesClassifierfromsklearn.decompositionimportPCAimportosfromsklearn.feature_selectionimportRFEfromsklearn.linear_modelimportLogisticRegression
- loading datasets
file=os.getcwd()+"/datasets_228_482_diabetes.csv"names= ['preg','plas','pres','skin','test','mass','pedi','age','class']df=pd.read_csv(file,names=names)array=df.valuesX=array[:,0:8]y=array[:,8]
- Different feature selection techniques:
SelectKBest
test=SelectKBest(score_func=f_classif,k=4)fit=test.fit(X,y)features=fit.transform(X)corr_p=df['skin'].corr(df['class'])print(corr_p)print(features[0:5,:])model=LogisticRegression(solver='lbfgs')rfe=RFE(model,3)fit=rfe.fit(X,y)print('Num features: %d'%fit.n_features_)print('Selected features: %s'%fit.support_)print('feature ranking: %s'%fit.ranking_)
ExtraTreeClasssifier
model=ExtraTreesClassifier(n_estimators=10)model.fit(X,y)print(model.feature_importances_)
Dimensionality Reduction- PCA
pca=PCA(n_components=3)fit=pca.fit(X,y)print('Explained Variance: %s'%fit.explained_variance_ratio_)print(fit.components_)
best params and score findings
lasso=Lasso()parameters= {'alpha': [1e-15,1e-10,1e-8,1e-4,1e-3,1e-2,1,5,10,20]}lasso_regressor=GridSearchCV(lasso,parameters,scoring='neg_mean_squared_error',cv=5)lasso_regressor.fit(X,y)print(lasso_regressor.best_params_)print(lasso_regressor.best_score_)
Connect-Linkedin
Website-RakibHHridoy
About
Before training a model or feed a model, first priority is on data,not in model. The more data is preprocessed and engineered the more model will learn. Feature selectio one of the methods processing data before feeding the model. Various feature selection techniques is shown here.
Topics
Resources
Uh oh!
There was an error while loading.Please reload this page.