[1]ZUO Pengyu,ZHOU Jie,WANG Shitong.Incremental online sequential extreme learning machine for imbalanced data[J].CAAI Transactions on Intelligent Systems,2020,15(3):520-527.[doi:10.11992/tis.201904040]
Copy
CAAI Transactions on Intelligent Systems[ISSN 1673-4785/CN 23-1538/TP] Volume:
15
Number of periods:
2020 3
Page number:
520-527
Column:
学术论文—人工智能基础
Public date:
2020-05-05
- Title:
-
Incremental online sequential extreme learning machine for imbalanced data
- Author(s):
-
ZUO Pengyu1; ZHOU Jie1; WANG Shitong1; 2
-
1. College of Digital Media, Jiangnan University, Wuxi 214122, China;
2. Jiangsu Province Key Lab. of Media Design & Software Technologies, Wuxi 214122, China
-
- Keywords:
-
class imbalance; incremental learning; inverse-free matrix; online learning; extreme learning machine; classification; multi-class imbalanced; neural network
- CLC:
-
TP181
- DOI:
-
10.11992/tis.201904040
- Abstract:
-
In this paper, an incremental online sequential extreme learning machine (IOS-ELM) is proposed to solve the problems of low efficiency and poor classification accuracy of OS-ELM for class imbalance learning. The basic idea is to adjust the balance factor according to the category imbalance ratio in an imbalanced dataset and then determine an optimal number of hidden nodes using the generalized inverse of the block matrix, thereby improving the online learning ability of IOS-ELM. The experiments on the effectiveness and feasibility of 14 binary-class and multi-class imbalanced datasets show that the proposed IOS-ELM has better generalization capability and classification performance than other comparative methods.