Abstract:Imbalanced data classification remains a research hotspot and a challenging problem in the field of machine learning. The challenge of imbalanced learning lies not only in class imbalance problem, but also in the class overlapping problem which is complex. However, most of the existing algorithms mainly focus on the former. The limitation prevents the existing methods from breaking through. To address this limitation, this paper proposes an ensemble algorithm based on dual clustering and stage-wise hybrid sampling (DCSHS) to address both class imbalance and class overlapping problems. The DCSHS has three main parts: projection clustering combination framework (PCC), stage-wise hybrid sampling (SHS) and envelope clustering transfer mapping mechanism (CTM). PCC is to create multiple subsets through projective clustering. SHS is to identify the overlapping region of each subset and conduct hybrid sampling. CTM is to explore more information of samples in each subset by combining the clustering and transfer learning. At first, we design a PCC framework guided by Davies-Bouldin clustering effectiveness index (DBI), which is used to obtain high-quality clusters and combine them to obtain a set of cross-complete subsets (CCS) with low overlapping. Secondly, according to the characteristics of subset classes, a SHS algorithm is designed to realize the de-overlapping and balancing of subsets. Finally, an envelope clustering transfer mapping mechanism (CTM) is constructed for all processed subsets by means of transfer learning, thereby reducing class overlapping and explore structural information of samples. Weak classifiers are trained on the balanced subsets, and fused as all the imbalanced ensemble algorithms did. The major advantage of our algorithm is that it can exploit the intersectionality of the CCS to realize the soft elimination of overlapping majority samples, and learn as much information of overlapping samples as possible, thereby enhancing the class overlapping while class balancing. In the experimental section, more than 30 public datasets and over ten representative algorithms are chosen for verification. The experimental results show that the DCSHS is significantly best in terms of anti-overlapping, Recall, F1-M, G-M, AUC, and diversity.

ECC + +: An algorithm family based on ensemble of classifier chains for classifying imbalanced multi-label data

Multi-class AdaBoost ELM

A novel bayesian network-based ensemble classifier chains for multi-label classification

CC k EL: Compensation-based correlated k -labelsets for classifying imbalanced multi-label data

Selective Ensemble of Classifier Chains.

Multi-label Classification: Dealing with Imbalance by Combining Labels

Learning Imbalanced Multi-class Data with Optimal Dichotomy Weights

Boosting label weighted extreme learning machine for classifying multi-label imbalanced data

A partition-based problem transformation algorithm for classifying imbalanced multi-label data

Progressive Hybrid Classifier Ensemble for Imbalanced Data

Rebalancing Multi-Label Class-Incremental Learning

A new classifier for imbalanced data with iterative learning process and ensemble operating process

To Combat Multiclass Imbalanced Problems by Aggregating Evolutionary Hierarchical Classifiers

An Imbalanced Ensemble Learning Method Based on Dual Clustering and Stage-Wise Hybrid Sampling

Class-Imbalanced Complementary-Label Learning via Weighted Loss

A Novel Multi-class Classification Architecture Combining Population-based Sampling and Multi-expert Classifier for Imbalanced Data.

Label Cluster Chains for Multi-Label Classification

Self-paced Ensemble for Highly Imbalanced Massive Data Classification

Equalization Ensemble for Large Scale Highly Imbalanced Data Classification

Dual Balanced Class-Incremental Learning With im-Softmax and Angular Rectification

A Three-phase Augmented Classifiers Chain Approach Based on Co-occurrence Analysis for Multi-Label Classification