A virtual sample generation method based on kernel density estimation and copula function for imbalanced classification

Qunxiong Zhu, Shixiong Wang, Zhongsheng Chen, Yanlin He, Yuan Xu

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

4 Citations (Scopus)

Abstract

In the case of imbalanced data, classification models often achieve low accuracy. To solve this problem, this paper proposes a virtual sample generation method based on kernel density estimation and copula function. The kernel density estimation is used to estimate the probability density of each dimension of data, and the joint probability density of the samples is constructed by the copula function. The validation experiments are carried out by applying the proposed method to a numerical simulation and a yeast classification problem. Simulation results show that the proposed method can generate high-quality virtual samples and significantly improve the recognition accuracy.

Original languageEnglish
Title of host publicationProceedings of 2019 IEEE 8th Data Driven Control and Learning Systems Conference, DDCLS 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages969-975
Number of pages7
ISBN (Electronic)9781728114545
DOIs
Publication statusPublished - May 2019
Externally publishedYes
Event8th IEEE Data Driven Control and Learning Systems Conference, DDCLS 2019 - Dali, China
Duration: 24 May 201927 May 2019

Publication series

NameProceedings of 2019 IEEE 8th Data Driven Control and Learning Systems Conference, DDCLS 2019

Conference

Conference8th IEEE Data Driven Control and Learning Systems Conference, DDCLS 2019
Country/TerritoryChina
CityDali
Period24/05/1927/05/19

Keywords

  • Classification
  • Copula
  • Imbalance dataset
  • Kernel density estimation
  • Virtual sample generation

Fingerprint

Dive into the research topics of 'A virtual sample generation method based on kernel density estimation and copula function for imbalanced classification'. Together they form a unique fingerprint.

Cite this