TY - JOUR
T1 - A General Framework for Dimensionality Reduction of K-Means Clustering
AU - Wu, Tong
AU - Xiao, Yanni
AU - Guo, Muhan
AU - Nie, Feiping
N1 - Publisher Copyright:
© 2019, The Classification Society.
PY - 2020/10
Y1 - 2020/10
N2 - Dimensionality reduction plays an important role in many machine learning and pattern recognition applications. Linear discriminant analysis (LDA) is the most popular supervised dimensionality reduction technique which searches for the projection matrix that makes the data points of different classes to be far from each other while requiring data points of the same class to be close to each other. In this paper, trace ratio LDA is combined with K-means clustering into a unified framework, in which K-means clustering is employed to generate class labels for unlabeled data and LDA is used to investigate low-dimensional representation of data. Therefore, by combining the subspace clustering with dimensionality reduction together, the optimal subspace can be obtained. Differing from other existing dimensionality reduction methods, our novel framework is suitable for different scenarios: supervised, semi-supervised, and unsupervised dimensionality reduction cases. Experimental results on benchmark datasets validate the effectiveness and superiority of our algorithm compared with other relevant techniques.
AB - Dimensionality reduction plays an important role in many machine learning and pattern recognition applications. Linear discriminant analysis (LDA) is the most popular supervised dimensionality reduction technique which searches for the projection matrix that makes the data points of different classes to be far from each other while requiring data points of the same class to be close to each other. In this paper, trace ratio LDA is combined with K-means clustering into a unified framework, in which K-means clustering is employed to generate class labels for unlabeled data and LDA is used to investigate low-dimensional representation of data. Therefore, by combining the subspace clustering with dimensionality reduction together, the optimal subspace can be obtained. Differing from other existing dimensionality reduction methods, our novel framework is suitable for different scenarios: supervised, semi-supervised, and unsupervised dimensionality reduction cases. Experimental results on benchmark datasets validate the effectiveness and superiority of our algorithm compared with other relevant techniques.
KW - Dimensionality reduction
KW - K-means clustering
KW - Trace ratio LDA
UR - http://www.scopus.com/inward/record.url?scp=85071334375&partnerID=8YFLogxK
U2 - 10.1007/s00357-019-09342-4
DO - 10.1007/s00357-019-09342-4
M3 - 文章
AN - SCOPUS:85071334375
SN - 0176-4268
VL - 37
SP - 616
EP - 631
JO - Journal of Classification
JF - Journal of Classification
IS - 3
ER -