Abstract:Minority oversampling is currently one of the most popular and effective methods for handling imbalanced data. However, oversampling that relies on the observations of the minority class to generate new samples is not applicable in the scenario of imbalanced data with extremely scarce minority samples, because the strongly underrepresented minority class does not contain enough information to support the oversampling process. Since some recent studies have exhibited the effectiveness of using majority information to bootstrap oversampling, the neglect of class overlap in the sampling process would increase the overlapping degree and complicate the decision boundary. To this end, this paper proposes a Mahalanobis distance and Local information based OverSampling (MLOS) for highly imbalanced class-overlapped data. MLOS first employs the majority density to guide the sample synthesis, with Mahalanobis distance to extract the majority probability contour. Then for each minority seed sample, to avoid the generation of overlapping samples, MLOS constrain the synthetic process by finding the auxiliary sample (in its 5 nearest neighbors) with similar probability density value to the seed. Finally, MLOS uses a pair-wise data cleaning process to improve the visibility of the decision boundary according to the probability density of synthetic samples. Comparative experiments conducted on 16 highly imbalanced class-overlapped datasets, using 17 different methods, demonstrates the superiority of our proposed method in terms of three popular evaluation metrics AUC , G - mean and Recall for imbalance classification. The source code of MLOS is available at https://github.com/ytyancp/MLOS .

Over-sampling Algorithm Based on Adaboost in Unbalanced Data Set

Imbalanced Data Sets Classification Method Based on Over-Sampling Technique

Imbalanced Data Classification Algorithm Based on Integrated Sampling and Ensemble Learning.

Over-sampling Algorithm Based on Preliminary Classification in Imbalanced Data Sets Learning

WOTBoost: Weighted Oversampling Technique in Boosting for imbalanced learning

An Over Sampling Method of Unbalanced Data Based on Ant Colony Clustering

An oversampling FCM-KSMOTE algorithm for imbalanced data classification

Over-sampling algorithm for imbalanced data classification

An Improving Majority Weighted Minority Oversampling Technique for Imbalanced Classification Problem

Adapting MultiBoost ensemble for class imbalanced learning

IMBALANCED DATA CLASSIFICATION ACTIVE LEARNING ALGORITHM BASED ON BOOSTING

A Classfication Method For Imbalance Data Set Based on Kernel SMOTE

A Novel Adaptive Minority Oversampling Technique for Improved Classification in Data Imbalanced Scenarios

A cluster-based SMOTE both-sampling (CSBBoost) ensemble algorithm for classifying imbalanced data

Hybrid SVM algorithm oriented to classifying imbalanced datasets

A majority affiliation based under-sampling method for class imbalance problem

Improved Pso_adaboost Ensemble Algorithm For Imbalanced Data

An Improved MAHAKIL Oversampling Method for Imbalanced Dataset Classification

Synthetic oversampling with Mahalanobis distance and local information for highly imbalanced class-overlapped data

Enhancing classification performance through multi-source online transfer learning algorithm with oversampling

Efficient hybrid oversampling and intelligent undersampling for imbalanced big data classification