An extended ISOMAP by enhancing similarity for clustering

Authors:
Hong Yu;Xianchao Zhang;Yuansheng Yang;Xiaowei Zhao;Lei Cai
Affiliations:
School of Electronics and Information Engineering, Dalian University of Technology, Dalian, China,Software School, Dalian University of Technology, Dalian, China;School of Electronics and Information Engineering, Dalian University of Technology, Dalian, China,Software School, Dalian University of Technology, Dalian, China;School of Electronics and Information Engineering, Dalian University of Technology, Dalian, China,Software School, Dalian University of Technology, Dalian, China;School of Electronics and Information Engineering, Dalian University of Technology, Dalian, China,Software School, Dalian University of Technology, Dalian, China;School of Electronics and Information Engineering, Dalian University of Technology, Dalian, China,Software School, Dalian University of Technology, Dalian, China
Venue:
IEA/AIE'12 Proceedings of the 25th international conference on Industrial Engineering and Other Applications of Applied Intelligent Systems: advanced research in applied artificial intelligence
Year:
2012

Citing 5
Cited 0

Algorithms for clustering data

Algorithms for clustering data
Cluster ensembles --- a knowledge reuse framework for combining multiple partitions

The Journal of Machine Learning Research
Pattern Classification (2nd Edition)

Pattern Classification (2nd Edition)
Flexible manifold embedding: a framework for semi-supervised and unsupervised dimension reduction

IEEE Transactions on Image Processing
An incremental manifold learning algorithm based on the small world model

LSMS/ICSEE'10 Proceedings of the 2010 international conference on Life system modeling and and intelligent computing, and 2010 international conference on Intelligent computing for sustainable energy and environment: Part I

Quantified Score

Hi-index	0.00

Visualization

Abstract

Isomap is an important dimension reduction method for clustering data with relatively large features. Isomap uses geodesic distance instead of Euclidean distance to reflect geometry of the underlying manifold, while it ignores the classification principle that the distance between samples on different manifolds should be large and the distance between samples on the same manifold should be small. In this paper, we employed a path based distance to extend Isomap for clustering. The path based distance measure strengthens the similarity of the points on the same manifold. The useful behavior of the similarity strengthening Isomap is confirmed through numerical experiments with several data sets.