首页|Balancing large margin nearest neighbours for imbalanced data

Balancing large margin nearest neighbours for imbalanced data

扫码查看
原文链接
  • NETL
  • NSTL
  • Inst Engineering Technology-Iet
It is critical to learn and obtain a good distance metric that can precisely measure the distance between samples in imbalanced data. However, traditional metric learning algorithms, e.g. large margin nearest neighbour (LMNN), information-theoretic metric learning, neighbourhood component analysis, do not take imbalanced distributions of classes into consideration. The traditional methods are apt to be affected by the majority samples, so those important minority samples are often ignored during the learning phase of distance metrics matrix, this may gravely confuse decision-making systems on classifying samples. In order to resolve this problem, the authors propose a novel metric-learning method named balancing large margin nearest neighbour (BLMNN) for imbalanced data. BLMNN can improve the objective function according to the distribution of classes, which treats the minority and majority classes equally during the optimisation process. Thus, the contribution of minority class is taken into full consideration, which can greatly improve the accuracy of classification. Substantial experiments were performed on real-world imbalanced datasets. The experiments results in various evaluation indexes of the proposed method comparing it with other metric-learning methods show the advantages of the proposed method.

learning (artificial intelligence)pattern classificationnearest neighbour methodsimbalanced dataimportant minority samplesdistance metrics matrixnovel metric-learning methodreal-world imbalanced datasetsmetric-learning methodsmachine learning tasksdistance measurementsnearest neighboursdistance metricinformation-theoretic metric learningneighbourhood component analysisBLMNN

Zhang, Xiaotian、Han, Nan、Qiao, Shaojie、Zhang, Yongqing、Huang, Ping、Peng, Jing、Zhou, Kai、Yuan, Chang-an

展开 >

Chengdu Univ Informat Technol, Sch Software Engn, Chengdu 610225, Peoples R China

Chengdu Univ Informat Technol, Sch Management, Chengdu 610103, Peoples R China

Chengdu Univ Informat Technol, Sch Comp Sci, Chengdu 610225, Peoples R China

Sichuan Prov Dept Publ Secur, Chengdu 610014, Peoples R China

Nanning Normal Univ, Sch Comp & Informat Engn, Nanning 530299, Peoples R China

展开 >

2020

10.1049/joe.2019.1178

The Journal of Engineering

The Journal of Engineering

ISSN：
年,卷(期)：2020.2020(13)
  • 2
  • 16