Abstract
Class evolution, the phenomenon of class emergence and disappearance, is an important research topic for data stream mining. All previous studies implicitly regard class evolution as a transient change, which is not true for many real-world problems. This paper concerns the scenario where classes emerge or disappear gradually. A class-based ensemble approach, namely Class-Based ensemble for Class Evolution (CBCE), is proposed. By maintaining a base learner for each class and dynamically updating the base learners with new data, CBCE can rapidly adjust to class evolution. A novel under-sampling method for the base learners is also proposed to handle the dynamic class-imbalance problem caused by the gradual evolution of classes. Empirical studies demonstrate the effectiveness of CBCE in various class evolution scenarios in comparison to existing class evolution adaptation methods. © 1989-2012 IEEE.
| Original language | English |
|---|---|
| Article number | 7401075 |
| Pages (from-to) | 1532-1545 |
| Number of pages | 14 |
| Journal | IEEE Transactions on Knowledge and Data Engineering |
| Volume | 28 |
| Issue number | 6 |
| Early online date | 8 Feb 2016 |
| DOIs | |
| Publication status | Published - 1 Jun 2016 |
| Externally published | Yes |
Funding
This work was supported in part by the National Natural Science Foundation of China under Grant 61329302 and Grant 61175065, in part by the Program for New Century Excellent Talents in University under Grant NCET-12-0512, and in part by EPSRC grant (Grant No. EP/J017515/1). Xin Yao was also supported by a Royal Society Wolfson Research Merit Award. Ke Tang is the corresponding author of this paper.
Keywords
- class evolution
- data stream mining
- ensemble model
- imbalanced classification
- on-line learning