my dataset test is 0 17565 1 2435 train is 0 70212 1 9788 I applied oversampling Smote with IsolationForest algorithm on just training set before oversampling results: F1 Score : 0.9278732648748262 Accuracy Score : 0.93025 Classification Report : precision recall f1-score support
0 0.95 0.97 0.96 17648
1 0.74 0.63 0.68 2352
after oversampling results: F1 Score : 0.8172379348236138 Accuracy Score : 0.83115 Classification Report : precision recall f1-score support
0 0.89 0.93 0.91 17648
1 0.18 0.13 0.15 2352
my code is:
clf=IsolationForest(n_estimators=50, max_samples='auto',
contamination=float(0.1),max_features=1.0)
clf.fit(X_train)
y_pred = clf.predict(X_test)
y_pred[y_pred == 1] = 0
y_pred[y_pred == -1] = 1
These results are normal? Because ı think after oversampling resuts must be improved. Could you help me? Best regards.
