Linear SVC Sklearn - Training a Linear SVM Classification Model
Python
An example of training a linear SVM classification model using SVC from sklearn. Here we set the C parameter to 1, the kernel to 'linear' and the class_weight to 'balanced'.
If the model is overfitting then reduce the C value.
The balanced class weight parameter addresses issues with imbalanced training data.
1| from sklearn.svm import SVC 2| from sklearn.metrics import classification_report 3| 4| # create a linear SVC model with balanced class weights 5| model = SVC(C=1, kernel='linear', class_weight='balanced') 6| 7| # fit model 8| model.fit(X_train, y_train) 9| 10| # make predictions on test data 11| y_pred = model.predict(X_test) 12| 13| # create a dataframe of feature coefficients 14| coef = pd.DataFrame(model.coef_,columns=X_train.columns) 15| print(coef) 16| 17| # print classification report 18| print(classification_report(y_test, y_pred))
149
132
127
119