SMOTE-ENC: A Novel SMOTE-Based Method to Generate Synthetic Data for Nominal and Continuous Features

Real-world datasets are heavily skewed where some classes significantly outnumbered by the other classes. In these situations, machine learning algorithms fail to achieve substantial efficacy while predicting underrepresented instances. To solve this problem, many variations of synthetic minority oversampling methods (SMOTE) have been proposed balance which deal with continuous features. However, for both nominal and features, SMOTE-NC is only SMOTE-based technique data. paper, we present a novel method, SMOTE-ENC (SMOTE—Encoded Nominal Continuous), in features encoded as numeric values difference between two such reflects amount change association class. Our experiments show that classification models using method offer better prediction than when dataset has number also there categorical target Additionally, our addressed one major limitations algorithm. can be applied on mixed consisting cannot function if all nominal. generalized nominal-only datasets.